Articoli di riviste: "Encoder optimization"

1

Hassan, Hammad, Muhammad Nasir Khan, Syed Omer Gilani, Mohsin Jamil, Hasan Maqbool, Abdul Waheed Malik e Ishtiaq Ahmad. "H.264 Encoder Parameter Optimization for Encoded Wireless Multimedia Transmissions". IEEE Access 6 (2018): 22046–53. http://dx.doi.org/10.1109/access.2018.2824835.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

2

Hamza, Ahmed M., Mohamed Abdelazim, Abdelrahman Abdelazim e Djamel Ait-Boudaoud. "HEVC Rate-Distortion Optimization with Source Modeling". Electronic Imaging 2021, n. 10 (18 gennaio 2021): 259–1. http://dx.doi.org/10.2352/issn.2470-1173.2021.10.ipas-259.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

The Rate-Distortion adaptive mechanisms of MPEG-HEVC (High Efficiency Video Coding) and its derivatives are an incremental improvement in the software reference encoder, providing a selective Lagrangian parameter choice which varies by encoding mode (intra or inter) and picture reference level. Since this weighting factor (and the balanced cost functions it impacts) are crucial to the RD optimization process, affecting several encoder decisions and both coding efficiency and quality of the encoded stream, we investigate an improvement by modern reinforcement learning methods. We develop a neural-based agent that learns a real-valued control policy to maximize rate savings by input signal pattern, mapping pixel intensity values from the picture at the coding tree unit level, to the appropriate weighting-parameter. Our testing on reference software yields improvements for coding efficiency performance across different video sequences, in multiple classes of video.

3

Wang, Lei, Qimin Ren, Jingang Jiang, Hongxin Zhang e Yongde Zhang. "Recent Patents on Magnetic Encoder and its use in Rotating Mechanism". Recent Patents on Engineering 13, n. 3 (19 settembre 2019): 194–200. http://dx.doi.org/10.2174/1872212112666180628145856.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

Background: The application of magnetic encoder relieves the problem of reliable application of servo system in vibration field. The magnetic encoder raises the efficiency and reliability of the system, and from structural considerations, the magnetic encoder is divided into two parts: signal conversion and structural support. Objective: In order to improve the accuracy of the magnetic encoder, its structure is constantly improving. To evaluate a magnetic encoder, the accuracy is a factor, meanwhile, the structure of magnetic encoder is one of the key factors that make difference in the accuracy of magnetic encoder. The purpose of this paper is to study the accuracy of different structures of magnetic encoder. Methods: This paper reviews various representative patents related to magnetic encoder. Results: The differences in different types of magnetic encoders were compared and analyzed and the characteristics were concluded. The main problems in its development were analyzed, the development trend forecasted, and the current and future developments of the patents on magnetic encoder were discussed. Conclusion: The optimization of the magnetic encoder structure improves the accuracy of magnetic encoder. In the future, for wide popularization of magnetic encoder, modularization, generalization, and reliability are the factors that practitioner should pay attention to, and more patents on magnetic encoder should be invented.

4

Lee, Yoon Jin, Dong In Bae e Gwang Hoon Park. "HEVC Encoder Optimization using Depth Information". Journal of Broadcast Engineering 19, n. 5 (30 settembre 2014): 640–55. http://dx.doi.org/10.5909/jbe.2014.19.5.640.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

5

Wang, Shanshe, Falei Luo, Siwei Ma, Xiang Zhang, Shiqi Wang, Debin Zhao e Wen Gao. "Low complexity encoder optimization for HEVC". Journal of Visual Communication and Image Representation 35 (febbraio 2016): 120–31. http://dx.doi.org/10.1016/j.jvcir.2015.12.005.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

6

Merel, Josh, Donald M. Pianto, John P. Cunningham e Liam Paninski. "Encoder-Decoder Optimization for Brain-Computer Interfaces". PLOS Computational Biology 11, n. 6 (1 giugno 2015): e1004288. http://dx.doi.org/10.1371/journal.pcbi.1004288.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

7

Hanli Wang, Ming-Yan Chan, S. Kwong e Chi-Wah Kok. "Novel quantized DCT for video encoder optimization". IEEE Signal Processing Letters 13, n. 4 (aprile 2006): 205–8. http://dx.doi.org/10.1109/lsp.2005.863691.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

8

Bariani, M., P. Lambruschini e M. Raggio. "An Efficient Multi-Core SIMD Implementation for H.264/AVC Encoder". VLSI Design 2012 (29 maggio 2012): 1–14. http://dx.doi.org/10.1155/2012/413747.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

The optimization process of a H.264/AVC encoder on three different architectures is presented. The architectures are multi- and singlecore and SIMD instruction sets have different vector registers size. The need of code optimization is fundamental when addressing HD resolutions with real-time constraints. The encoder is subdivided in functional modules in order to better understand where the optimization is a key factor and to evaluate in details the performance improvement. Common issues in both partitioning a video encoder into parallel architectures and SIMD optimization are described, and author solutions are presented for all the architectures. Besides showing efficient video encoder implementations, one of the main purposes of this paper is to discuss how the characteristics of different architectures and different set of SIMD instructions can impact on the target application performance. Results about the achieved speedup are provided in order to compare the different implementations and evaluate the more suitable solutions for present and next generation video-coding algorithms.

9

Cho, Jung-Hyun, Myung-Soo Lee, Han-Soo Jeong, Chang-Suk Kim e Dae-Jea Cho. "Optimization of H.264 Encoder based on Hardware Implementation in Embedded System". Journal of the Korea Academia-Industrial cooperation Society 11, n. 8 (31 agosto 2010): 3076–82. http://dx.doi.org/10.5762/kais.2010.11.8.3076.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

10

Hou, Han, Guohua Cao, Hongchang Ding e Kun Li. "Research on Particle Swarm Compensation Method for Subdivision Error Optimization of Photoelectric Encoder Based on Parallel Iteration". Sensors 22, n. 12 (12 giugno 2022): 4456. http://dx.doi.org/10.3390/s22124456.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

Photoelectric encoders are widely used in high-precision measurement fields such as industry and aerospace because of their high precision and reliability. In order to improve the subdivision accuracy of moiré grating signals, a particle swarm optimization compensation model for grating the subdivision error of a photoelectric encoder based on parallel iteration is proposed. In the paper, an adaptive subdivision method of a particle swarm search domain based on the honeycomb structure is proposed, and a raster signal subdivision error compensation model based on the multi-swarm particle swarm optimization algorithm based on parallel iteration is established. The optimization algorithm can effectively improve the convergence speed and system accuracy of traditional particle swarm optimization. Finally, according to the subdivision error compensation algorithm, the subdivision error of the grating system caused by the sinusoidal error in the system is quickly corrected by taking advantage of the high-speed parallel processing of the FPGA pipeline architecture. The design experiment uses a 25-bit photoelectric encoder to verify the subdivision error algorithm. The experimental results show that the actual dynamic subdivision error can be reduced to ½ before compensation, and the static subdivision error can be reduced from 1.264″ to 0.487″ before detection.

11

Karim, Ahmad M., Hilal Kaya, Mehmet Serdar Güzel, Mehmet R. Tolun, Fatih V. Çelebi e Alok Mishra. "A Novel Framework Using Deep Auto-Encoders Based Linear Model for Data Classification". Sensors 20, n. 21 (9 novembre 2020): 6378. http://dx.doi.org/10.3390/s20216378.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

This paper proposes a novel data classification framework, combining sparse auto-encoders (SAEs) and a post-processing system consisting of a linear system model relying on Particle Swarm Optimization (PSO) algorithm. All the sensitive and high-level features are extracted by using the first auto-encoder which is wired to the second auto-encoder, followed by a Softmax function layer to classify the extracted features obtained from the second layer. The two auto-encoders and the Softmax classifier are stacked in order to be trained in a supervised approach using the well-known backpropagation algorithm to enhance the performance of the neural network. Afterwards, the linear model transforms the calculated output of the deep stacked sparse auto-encoder to a value close to the anticipated output. This simple transformation increases the overall data classification performance of the stacked sparse auto-encoder architecture. The PSO algorithm allows the estimation of the parameters of the linear model in a metaheuristic policy. The proposed framework is validated by using three public datasets, which present promising results when compared with the current literature. Furthermore, the framework can be applied to any data classification problem by considering minor updates such as altering some parameters including input features, hidden neurons and output classes.

12

Yang, Xiaoshan, Baochen Xiong, Yi Huang e Changsheng Xu. "Cross-Modal Federated Human Activity Recognition via Modality-Agnostic and Modality-Specific Representation Learning". Proceedings of the AAAI Conference on Artificial Intelligence 36, n. 3 (28 giugno 2022): 3063–71. http://dx.doi.org/10.1609/aaai.v36i3.20213.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

In this paper, we propose a new task of cross-modal federated human activity recognition (CMF-HAR), which is conducive to promote the large-scale use of the HAR model on more local devices. To address the new task, we propose a feature-disentangled activity recognition network (FDARN), which has five important modules of altruistic encoder, egocentric encoder, shared activity classifier, private activity classifier and modality discriminator. The altruistic encoder aims to collaboratively embed local instances on different clients into a modality-agnostic feature subspace. The egocentric encoder aims to produce modality-specific features that cannot be shared across clients with different modalities. The modality discriminator is used to adversarially guide the parameter learning of the altruistic and egocentric encoders. Through decentralized optimization with a spherical modality discriminative loss, our model can not only generalize well across different clients by leveraging the modality-agnostic features but also capture the modality-specific discriminative characteristics of each client. Extensive experiment results on four datasets demonstrate the effectiveness of our method.

13

Cahani, Ilda, e Marcus Stiemer. "Mathematical optimization and machine learning to support PCB topology identification". Advances in Radio Science 21 (1 dicembre 2023): 25–35. http://dx.doi.org/10.5194/ars-21-25-2023.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

Abstract. In this paper, we study an identification problem for schematics with different concurring topologies. A framework is proposed, that is both supported by mathematical optimization and machine learning algorithms. Through the use of Python libraries, such as scikit-rf, which allows for the emulation of network analyzer measurements, and a physical microstrip line simulation on PCBs, data for training and testing the framework are provided. In addition to an individual treatment of the concurring topologies and subsequent comparison, a method is introduced to tackle the identification of the optimum topology directly via a standard optimization or machine learning setup: An encoder-decoder sequence is trained with schematics of different topologies, to generate a flattened representation of the rated graph representation of the considered schematics. Still containing the relevant topology information in encoded (i.e., flattened) form, the so obtained latent space representations of schematics can be used for standard optimization of machine learning processes. Using now the encoder to map schematics on latent variables or the decoder to reconstruct schematics from their latent space representation, various machine learning and optimization setups can be applied to treat the given identification task. The proposed framework is presented and validated for a small model problem comprising different circuit topologies.

14

Wali, Ibtissem, Amina Kessentini, Mohamed Ali Ben Ayed e Nouri Masmoudi. "DSP TMS320C6678 Based SHVC Encoder Implementation and its Optimization". International Journal of Recent Technology and Engineering (IJRTE) 10, n. 5 (30 gennaio 2022): 24–31. http://dx.doi.org/10.35940/ijrte.e6656.0110522.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

The programmable processors newest technologies, as for example the multicore Digital Signal Processors (DSP), offer a promising solution for overcoming the complexity of the real time video encoding application. In this paper, the SHVC video encoder was effectively implemented just on a single core among the eight cores of TMS320C6678 DSP for a Common Intermediate Format (CIF)input video sequence resolution(352x288). Performance optimization of the SHVC encoder had reached up 41% compared to its reference software enabling a real-time implementation of the SHVC encoder for CIF input videos sequence resolution. The proposed SHVC implementation was carried out on different quantization parameters (QP). Several experimental tests had proved our performance achievement for real-time encoding on TMS320C6678.

15

Bitriá, Ricard, e Jordi Palacín. "Optimal PID Control of a Brushed DC Motor with an Embedded Low-Cost Magnetic Quadrature Encoder for Improved Step Overshoot and Undershoot Responses in a Mobile Robot Application". Sensors 22, n. 20 (14 ottobre 2022): 7817. http://dx.doi.org/10.3390/s22207817.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

The development of a proportional–integral–derivative (PID) control system is a simple, practical, highly effective method used to control the angular rotational velocity of electric motors. This paper describes the optimization of the PID control of a brushed DC motor (BDCM) with an embedded low-cost magnetic quadrature encoder. This paper demonstrates empirically that the feedback provided by low-cost magnetic encoders produces some inaccuracies and control artifacts that are not usually considered in simulations, proposing a practical optimization approach in order to improve the step overshoot and undershoot controller response. This optimization approach is responsible for the motion performances of a human-sized omnidirectional mobile robot using three motorized omnidirectional wheels.

16

Hung-Chi Fang, Yu-Wei Chang, Tu-Chih Wang, Chao-Tsung Huang e Liang-Gee Chen. "High-performance JPEG 2000 encoder with rate-distortion optimization". IEEE Transactions on Multimedia 8, n. 4 (agosto 2006): 645–53. http://dx.doi.org/10.1109/tmm.2006.876305.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

17

Vudadha, Chetan, Srinivasan Rajagopalan, Aditya Dusi, P. Sai Phaneendra e M. B. Srinivas. "Encoder-Based Optimization of CNFET-Based Ternary Logic Circuits". IEEE Transactions on Nanotechnology 17, n. 2 (marzo 2018): 299–310. http://dx.doi.org/10.1109/tnano.2018.2800015.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

18

Dumitrescu, Sorina. "Fast Encoder Optimization for Multi-Resolution Scalar Quantizer Design". IEEE Transactions on Information Theory 57, n. 3 (marzo 2011): 1520–29. http://dx.doi.org/10.1109/tit.2011.2104990.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

19

Gurauskis, Donatas, Krzysztof Przystupa, Artūras Kilikevičius, Mikołaj Skowron, Matijošius Jonas, Joanna Michałowska e Kristina Kilikevičienė. "Performance Analysis of an Experimental Linear Encoder’s Reading Head under Different Mounting and Dynamic Conditions". Energies 15, n. 16 (22 agosto 2022): 6088. http://dx.doi.org/10.3390/en15166088.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

The performance of an optical linear encoder is described and evaluated by certain parameters such as its resolution, accuracy and repeatability. The best encoder for a particular application, just like other sensors, is usually selected according these parameters. There are, however, many side effects that have a direct influence on the optimal operation of an encoder. In order to understand how to minimize these harmful effects, a deeper knowledge of an encoder’s performance and a method for determining these factors are necessary. The main aspects of an encoder’s accuracy, resolution and repeatability are briefly reviewed in this paper. Discussed and developed in previous work, the experimental reading head for a Moiré effect-based optical linear encoder is used for the experimental analysis of the influence of different reading head designs on an encoder’s performance under various mounting inaccuracies and dynamic conditions.

20

Scanlan, Anthony, Daniel O’Hare, Mark Halton, Vincent O’Brien, Brendan Mullane e Eric Thompson. "Analysis of feedback predictive encoder based ADCs". COMPEL - The international journal for computation and mathematics in electrical and electronic engineering 36, n. 1 (3 gennaio 2017): 129–52. http://dx.doi.org/10.1108/compel-12-2015-0464.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

Purpose The purpose of this paper is to present analysis of the feedback predictive encoder-based analog-to-digital converter (ADC). Design/methodology/approach The use of feedback predictive encoder-based ADCs presents an alternative to the traditional two-stage pipeline ADC by replacing the input estimate producing first stage of the pipeline with a predictive loop that also produces an estimate of the input signal. Findings The overload condition for feedback predictive encoder ADCs is dependent on input signal amplitude and frequency, system gain and filter order. The limitation on the practical usable filter order is set by limit cycle oscillation. A boundary condition is defined for determination of maximum usable filter order. In a practical implementation of the predictive encoder ADC, the time allocated to the key functions of the gain stage and loop quantizer leads to optimization of the power consumption. Practical implications A practical switched capacitor implementation of the predictive encoder-based ADC is proposed. The power consumption of key circuit blocks is investigated. Originality/value This paper presents a methodology to optimize the bandwidth of predictive encoder ADCs. The overload and stability conditions may be used to determine the maximum input signal bandwidth for a given loop quantizer. Optimization of power consumption based on the allocation of time between the gain stage and the successive approximation register ADC operation is investigated. The lower bound of power consumption for this architecture is estimated.

21

Chen, Baifan, Haowu Zhao, Ruyi Zhu e Yemin Hu. "Marked-LIEO: Visual Marker-Aided LiDAR/IMU/Encoder Integrated Odometry". Sensors 22, n. 13 (23 giugno 2022): 4749. http://dx.doi.org/10.3390/s22134749.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

In this paper, we propose a visual marker-aided LiDAR/IMU/encoder integrated odometry, Marked-LIEO, to achieve pose estimation of mobile robots in an indoor long corridor environment. In the first stage, we design the pre-integration model of encoder and IMU respectively to realize the pose estimation combined with the pose estimation from the second stage providing prediction for the LiDAR odometry. In the second stage, we design low-frequency visual marker odometry, which is optimized jointly with LiDAR odometry to obtain the final pose estimation. In view of the wheel slipping and LiDAR degradation problems, we design an algorithm that can make the optimization weight of encoder odometry and LiDAR odometry adjust adaptively according to yaw angle and LiDAR degradation distance respectively. Finally, we realize the multi-sensor fusion localization through joint optimization of an encoder, IMU, LiDAR, and camera measurement information. Aiming at the problems of GNSS information loss and LiDAR degradation in indoor corridor environment, this method introduces the state prediction information of encoder and IMU and the absolute observation information of visual marker to achieve the accurate pose of indoor corridor environment, which has been verified by experiments in Gazebo simulation environment and real environment.

22

Wang, Xiaoyi, Chengxiang Zhao, Longyuan Xiao, Kunlei Zheng, Mingkang Liu, Dongjie Zhu e Tianyang Yao. "Self-Calibration Method for Circular Encoder Based on Two Reading Heads with Adjustable Positions". Machines 12, n. 4 (9 aprile 2024): 246. http://dx.doi.org/10.3390/machines12040246.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

To improve the measuring accuracy of circular encoders in special applications, self-calibration of the circular encoder is necessary. The commonly used self-calibration method, namely the Equal Division Average (EDA) method, requires a large number of reading heads and cannot be used when the structural space is limited. In this paper, a self-calibration method for a circular encoder based on two reading heads with adjustable positions (TRAP) is proposed. This TRAP method uses two reading heads to simulate multiple reading heads, which can be used to achieve self-calibration of circular encoders in limited space. This paper investigates the principle of simulating multiple reading heads with two reading heads, designs and builds an experimental system, and obtains and analyzes experimental data. The experimental results show that the peak-valley value of the angle measurement error is reduced from 252.41″ to 18.82″ after self-calibration with the TRAP method, and the repeatability of multiple self-calibration experimental results is less than 0.75″. The TRAP method breaks through the limitation of the number of reading heads that can be installed during the self-calibration of the circular encoder and can effectively suppress the angle measurement error of the circular encoder.

23

Huang, Yatao, Zihan Su, Di Chang, Yunke Sun e Jiubin Tan. "Error Analysis of an Economical On-Site Calibration System for Linear Optical Encoders". Metrology 4, n. 1 (13 marzo 2024): 131–40. http://dx.doi.org/10.3390/metrology4010009.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

A calibration system was designed to evaluate the accuracy of linear optical encoders at the micron level in a fast and economical manner. The system uses a commercial interferometer and motor stage as the calibrator and moving platform. Error analysis is necessary to prove the effectiveness and identify areas for optimization. A fixture was designed for the scale and interferometer target to meet the Abbe principle. A five-degree-of-freedom manual stage was utilized to adjust the reading head in optimal or suboptimal working conditions, such as working distance, offset, and angular misalignment. The results indicate that the calibration system has an accuracy of ±2.2 μm. The geometric errors of the calibration system, including mounting errors and non-ideal motions, are analyzed in detail. The system could be an inexpensive solution for encoder manufacturers and customers to calibrate a linear optical encoder or test its performance.

24

Madiwa, Shweta M., e Vishwanath Burkpalli. "Sine Cosine Based Harris Hawks Optimizer: A Hybrid Optimization Algorithm for Skin Cancer Detection Using Deep Stack Auto Encoder". Revue d'Intelligence Artificielle 36, n. 5 (23 dicembre 2022): 697–708. http://dx.doi.org/10.18280/ria.360506.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

Skin cancer is becoming major problems due to its tremendous growth. Skin cancer is a malignant skin lesion, which may cause damage to human. Hence, prior detection and precise medical diagnosis of the skin lesion is essential. In medical practice, detection of malignant lesions needs pathological examination and biopsy, which is expensive. The existing techniques need a brief physical inspection, which is imprecise and time-consuming. This paper presents a computer-assisted skin cancer detection strategy for detecting the skin lesion in skin images using deep stacked auto encoder. Sine Cosine-based Harris Hawks Optimizer (SCHHO) trains deep stacked auto encoders. The proposed SCHHO algorithm is designed by combining Sine Cosine Algorithm (SCA) and Harris Hawks Optimizer (HHO). The identification of skin lesion is performed on each segment, which is obtained by sparse-Fuzzy-c-means (FCM) algorithm. Statistical features, texture features and entropy are employed for selecting the most significant feature. Mean, standard deviation, variance, kurtosis, entropy, and Linear Discriminant Analysis (LDP) featured are extracted. SCHHO-Deep stacked auto-encoder outperformed other approaches with 91.66% accuracy, 91.60% sensitivity, and 91.72% specificity.

25

Yu, Junchi, Tingyang Xu, Yu Rong, Junzhou Huang e Ran He. "Structure-aware conditional variational auto-encoder for constrained molecule optimization". Pattern Recognition 126 (giugno 2022): 108581. http://dx.doi.org/10.1016/j.patcog.2022.108581.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

26

Chunlian, Yao, Liu Wen, Wu hongli, Mao Dianhui e Liu Li. "Bit Rate Buffer Control and Optimization of Embedded Video Encoder". Open Cybernetics & Systemics Journal 8, n. 1 (31 dicembre 2014): 1009–14. http://dx.doi.org/10.2174/1874110x01408011009.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

27

Xu, Long, Sam Kwong, Yun Zhang e Debin Zhao. "Low-Complexity Encoder Framework for Window-Level Rate Control Optimization". IEEE Transactions on Industrial Electronics 60, n. 5 (maggio 2013): 1850–58. http://dx.doi.org/10.1109/tie.2012.2190960.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

28

Gao, Bingshu, Dongdong Yu, Gong Chen e Wei Lu. "Online Calibration Based on Laser-Encoder Odometry". Journal of Physics: Conference Series 2520, n. 1 (1 giugno 2023): 012041. http://dx.doi.org/10.1088/1742-6596/2520/1/012041.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

Abstract This paper proposes a new online calibration method for the differential-drive mobile robot equipped with a laser scanner. Our algorithm can jointly estimate the 3-DoF extrinsic parameters of the laser scanner and the intrinsic parameters (radii and wheel spacing) of the differential-drive kinematic model. Applying the pre-integration theory initially developed the for IMU sensor to the differential-drive kinematic model, we adopt iterative nonlinear optimization to minimize the cost derived from laser and encoder measurements. Experiments results confirm that the proposed method can do online calibration precisely.

29

Whit Ney, Heh, Ab Al-Hadi Ab Rahman, Ainy Haziyah Awab, Mohd Shahrizal Rusli, Usman Ullah Sheikh e Goh Kam Meng. "Hardware design of a scalable and fast 2-D hadamard transform for HEVC video encoder". Indonesian Journal of Electrical Engineering and Computer Science 15, n. 3 (1 settembre 2019): 1401. http://dx.doi.org/10.11591/ijeecs.v15.i3.pp1401-1410.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

<span>This paper presents the hardware design of a 2-dimensional Hadamard transform used the in the rate distortion optimization module in state-of-the-art HEVC video encoder. The transform is mainly used to quickly determine optimum block size for encoding part of a video frame. The proposed design is both scalable and fast by 1) implementing a unified architecture for sizes 4x4 to 32x32, and 2) pipelining and feed through control that allows high performance for all block sizes. The design starts with high-level algorithmic loop unrolling optimization to determine suitable level of parallelism. Based on this, a suitable hardware architecture is devised using transpose memory buffer as pipeline memory for maximum performance. The design is synthesized and implemented on Xilinx Kintex Ultrascale FPGA. Results indicate variable performance obtained for different block sizes and higher operating frequency compared to a similar work in literature. The proposed design can be used as a hardware accelerator to speed up the rate distortion optimization operation in HEVC video encoders.</span>

30

Liu, Hongjian. "Research on Literary Translation Based on the Improved Optimization Model". Discrete Dynamics in Nature and Society 2022 (16 aprile 2022): 1–7. http://dx.doi.org/10.1155/2022/1329632.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

Machine translation is widely used in people’s daily lives and production, occupying an important position. In order to improve the accuracy of the literary intelligent translation, research on literary intelligent translation is based on the improved optimization model. Based on semantic features, the semantic ontology optimization model including an encoder and a decoder is created by machine translation. In order to improve the accuracy of the intelligent translation literature of the semantic ontology optimization model, the conversion layer, including the forward neural network layer, residual connection layer, and normalization layer, is added between the encoder and decoder of the semantic ontology optimization model. An improved optimization model is established, and syntax conversion is realized by using the conversion layer, which completes the intelligent translation of literature. It is found that the BLEU value of using this method to translate literary sentences can reach 17.23 when the number of training steps is set as 8000, and the training time is low. The translation result has a low correlation misalignment rate, which can meet the user’s literary translation needs.

31

Yang, Xi, Jie Zhang, Han Fang, Chang Liu, Zehua Ma, Weiming Zhang e Nenghai Yu. "AutoStegaFont: Synthesizing Vector Fonts for Hiding Information in Documents". Proceedings of the AAAI Conference on Artificial Intelligence 37, n. 3 (26 giugno 2023): 3198–205. http://dx.doi.org/10.1609/aaai.v37i3.25425.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

Hiding information in text documents has been a hot topic recently, with the most typical schemes of utilizing fonts. By constructing several fonts with similar appearances, information can be effectively represented and embedded in documents. However, due to the unstructured characteristic, font vectors are more difficult to synthesize than font images. Existing methods mainly use handcrafted features to design the fonts manually, which is time-consuming and labor-intensive. Moreover, due to the diversity of fonts, handcrafted features are not generalizable to different fonts. Besides, in practice, since documents might be distorted through transmission, ensuring extractability under distortions is also an important requirement. Therefore, three requirements are imposed on vector font generation in this domain: automaticity, generalizability, and robustness. However, none of the existing methods can satisfy these requirements well and simultaneously. To satisfy the above requirements, we propose AutoStegaFont, an automatic vector font synthesis scheme for hiding information in documents. Specifically, we design a two-stage and dual-modality learning framework. In the first stage, we jointly train an encoder and a decoder to invisibly encode the font images with different information. To ensure robustness, we target designing a noise layer to work with the encoder and decoder during training. In the second stage, we employ a differentiable rasterizer to establish a connection between the image and the vector modality. Then, we design an optimization algorithm to convey the information from the encoded image to the corresponding vector. Thus the encoded font vectors can be automatically generated. Extensive experiments demonstrate the superior performance of our scheme in automatically synthesizing vector fonts for hiding information in documents, with robustness to distortions caused by low-resolution screenshots, printing, and photography. Besides, the proposed framework has better generalizability to fonts with diverse styles and languages.

32

Li, Yibiao. "Optimization of Artistic Image Segmentation Algorithm Based on Feed Forward Neural Network under Complex Background Environment". Journal of Environmental and Public Health 2022 (13 settembre 2022): 1–11. http://dx.doi.org/10.1155/2022/9454344.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

Based on the theory and application, this paper discusses the optimization of art image segmentation algorithm based on FFNN (Feed Forward Neural Network). In this paper, residual units are used in the corresponding stages of encoder and decoder, and feature information of several convolution layers in each convolution stage of encoder is extracted at the same time. And the feature pyramid module is used to extract multiscale features from the feature map of the last convolution stage in the encoder. Finally, pixel by pixel additions combine the previously mentioned feature information into the corresponding layer of the decoder. Additionally, an improved weight adaptive algorithm based on feature preservation is suggested in this paper, which addresses the issue that the conventional image segmentation algorithm is noise-sensitive. The adaptive connection weight mechanism is also introduced. The accuracy and recall rates of this optimization algorithm can both reach 96.574%, according to the results of 50% cross-validation. All the segmentation performance evaluation indexes of this algorithm are higher than the existing main algorithms. Moreover, the algorithm takes a short time, does not need too much manual intervention, and can effectively segment artistic images. The optimization algorithm in this paper has certain reference significance for the related research of artistic image segmentation.

33

Guo, Hao, Meichao Song, Zhen Ding, Chunzhi Yi e Feng Jiang. "Vision-Based Efficient Robotic Manipulation with a Dual-Streaming Compact Convolutional Transformer". Sensors 23, n. 1 (3 gennaio 2023): 515. http://dx.doi.org/10.3390/s23010515.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

Learning from visual observation for efficient robotic manipulation is a hitherto significant challenge in Reinforcement Learning (RL). Although the collocation of RL policies and convolution neural network (CNN) visual encoder achieves high efficiency and success rate, the method general performance for multi-tasks is still limited to the efficacy of the encoder. Meanwhile, the increasing cost of the encoder optimization for general performance could debilitate the efficiency advantage of the original policy. Building on the attention mechanism, we design a robotic manipulation method that significantly improves the policy general performance among multitasks with the lite Transformer based visual encoder, unsupervised learning, and data augmentation. The encoder of our method could achieve the performance of the original Transformer with much less data, ensuring efficiency in the training process and intensifying the general multi-task performances. Furthermore, we experimentally demonstrate that the master view outperforms the other alternative third-person views in the general robotic manipulation tasks when combining the third-person and egocentric views to assimilate global and local visual information. After extensively experimenting with the tasks from the OpenAI Gym Fetch environment, especially in the Push task, our method succeeds in 92% versus baselines that of 65%, 78% for the CNN encoder, 81% for the ViT encoder, and with fewer training steps.

34

Wang, Lei, Xin Wei, Pengbo Liang, Yongde Zhang e Shuanghui Hao. "A Novel Angle Segmentation Method for Magnetic Encoders Based on Filtering Window Adaptive Adjustment Using Improved Particle Swarm Optimization". Sensors 23, n. 21 (25 ottobre 2023): 8695. http://dx.doi.org/10.3390/s23218695.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

In this paper we outline newly-developed control algorithms, designed to achieve high-precision feedback for a motor control system using a magnetic encoder. The magnetic encoder, combing single-pole and multi-pole magnetic steels, was adopted to extend the resolution of the magnetic encoder. First, with a view to settling the issue of the jump points of the multi-pole angle value at the convergence of two adjacent magnetic poles, the angle segmentation method, which uses the window filter discrimination method, is employed to determine the actual angle value. The appropriate filter window width is selected via the improved particle swarm optimization (IPSO) algorithm, and an expanded resolution is achieved. Second, a compensation table is completed via a linear compensation algorithm based on virtual cutting to enhance the accuracy of the combined magnetic encoder. On this basis, a linear difference algorithm is used to achieve deviation correction of the angle. Finally, the jump points can be restrained effectively via the angle segmentation method. The resolution reaches 0.05°, and the accuracy is 0.045°.

35

Yikui, He. "Optimization and Implementation of H.264 Encoder Based on DSP Platform". Journal of Physics: Conference Series 1982, n. 1 (1 luglio 2021): 012177. http://dx.doi.org/10.1088/1742-6596/1982/1/012177.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

36

You, Lei, Zhimin Gao, Yu Tong Han e Xin Su. "Transmission Optimization of Wireless Visual Sensor Networks with Compressed Sensing Encoder". Applied Mechanics and Materials 263-266 (dicembre 2012): 878–81. http://dx.doi.org/10.4028/www.scientific.net/amm.263-266.878.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

The design and optimization of wireless visual sensor network (WVSN) is challenging due to error-prone wireless link, limited resource and low computational capacity. Traditional video coding methods and layered transmission protocols are not efficient for visual information collection in WVSNs. In this paper, we adopted the compressed sensing (CS) paradigm to sense and encode the visual information, and proposed a cross-layer transmission algorithm to achieve optimal trade-off between total distortion and power consumption by simultaneously controlling the CS measurement rate and allocating the power and bandwidth for each wireless link. The algorithm was obtained by using convex dual theory and sub-gradient iterative method. The optimality condition of the proposed algorithm was also given. The algorithm can be implemented in a practical WVSN in a distributed way with only limited local information exchanges.

37

Lim, Duk-Gyu, Sharad Shakya e Je-Hoon Lee. "Optimization of a Systolic Array BCH encoder with Tree-Type Structure". International Journal of Contents 9, n. 1 (28 marzo 2013): 33–37. http://dx.doi.org/10.5392/ijoc.2013.9.1.033.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

38

Abdulaziz AlArfaj, Abeer, e Hanan Ahmed Hosni Mahmoud. "A Moving Object Tracking Technique Using Few Frames with Feature Map Extraction and Feature Fusion". ISPRS International Journal of Geo-Information 11, n. 7 (7 luglio 2022): 379. http://dx.doi.org/10.3390/ijgi11070379.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

Moving object tracking techniques using machine and deep learning require large datasets for neural model training. New strategies need to be invented that utilize smaller data training sizes to realize the impact of large-sized datasets. However, current research does not balance the training data size and neural parameters, which creates the problem of inadequacy of the information provided by the low visual data content for parameter optimization. To enhance the performance of moving object tracking that appears in only a few frames, this research proposes a deep learning model using an abundant encoder–decoder (a high-resolution transformer (HRT) encoder–decoder). An HRT encoder–decoder employs feature map extraction that focuses on high resolution feature maps that are more representative of the moving object. In addition, we employ the proposed HRT encoder–decoder for feature map extraction and fusion to reimburse the few frames that have the visual information. Our extensive experiments on the Pascal DOC19 and MS-DS17 datasets have implied that the HRT encoder–decoder abundant model outperforms those of previous studies involving few frames that include moving objects.

39

Zhang, Duo Li, Chuan Jie Wang, Yu Kun Song, Gao Ming Du e Xian Wen Cheng. "Design and Memory Optimization of Motion Estimation Module for H.264/AVC Encoder". Advanced Materials Research 179-180 (gennaio 2011): 1350–55. http://dx.doi.org/10.4028/www.scientific.net/amr.179-180.1350.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

H.264/AVC standard has been widely used in video compression at various kinds of application domain. Motion estimation takes the most calculation workload of H.264/AVC encoder. Memory optimization has played an even more important role in encoder design. Firstly, dependency relation between motion vectors was analyzed and removed at a little cost of estimation accuracy decrement, and then a 3-stage macro-block level pipeline architecture was proposed to increase parallel process ability of motion estimation. Then an optimized memory organization strategy of reference frame data was put forward, aiming at avoiding row changing frequently in SDRAM access. Finally, based on the 3-stage pipeline structure, a shared cyclic search window memory was proposed: 1) data relativity between adjacent macro-block was analyzed, 2) and search window memory size was elaborated, 3) and then a slice based structure and the work process were discussed. Analysis and experiment result show that 50% of on chip memory resource and cycles for off chip SDRAM access can be saved. The whole design was implemented with Verilog HDL and integrated into a H.264 encoder, which can demo 1280*720@30 video successfully at frequency of 120MHz under a cyclone III FPGA development board.

40

Galdran, Adrian. "Polyp and Surgical Instrument Segmentation with Double Encoder-Decoder Networks". Nordic Machine Intelligence 1, n. 1 (1 novembre 2021): 5–7. http://dx.doi.org/10.5617/nmi.9107.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

This paper describes a solution for the MedAI competition, in which participants were required to segment both polyps and surgical instruments from endoscopic images. Our approach relies on a double encoder-decoder neural network which we have previously applied for polyp segmentation, but with a series of enhancements: a more powerful encoder architecture, an improved optimization procedure, and the post-processing of segmentations, based on tempered model ensembling. Experimental results show that our method produces segmentations that show a good agreement with manual delineations provided by medical experts.

41

Yin, Yijun, Wenzheng Xu, Lei Chen e Hao Wu. "CoT-UNet++: A medical image segmentation method based on contextual transformer and dense connection". Mathematical Biosciences and Engineering 20, n. 5 (2023): 8320–36. http://dx.doi.org/10.3934/mbe.2023364.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

<abstract> <p>Accurate depiction of individual teeth from CBCT images is a critical step in the diagnosis of oral diseases, and the traditional methods are very tedious and laborious, so automatic segmentation of individual teeth in CBCT images is important to assist physicians in diagnosis and treatment. TransUNet has achieved success in medical image segmentation tasks, which combines the advantages of Transformer and CNN. However, the skip connection taken by TransUNet leads to unnecessary restrictive fusion and also ignores the rich context between adjacent keys. To solve these problems, this paper proposes a context-transformed TransUNet++ (CoT-UNet++) architecture, which consists of a hybrid encoder, a dense connection, and a decoder. To be specific, a hybrid encoder is first used to obtain the contextual information between adjacent keys by CoTNet and the global context encoded by Transformer. Then the decoder upsamples the encoded features by cascading upsamplers to recover the original resolution. Finally, the multi-scale fusion between the encoded and decoded features at different levels is performed by dense concatenation to obtain more accurate location information. In addition, we employ a weighted loss function consisting of focal, dice, and cross-entropy to reduce the training error and achieve pixel-level optimization. Experimental results demonstrate that the proposed CoT-UNet++ method outperforms the baseline models and can obtain better performance in tooth segmentation.</p> </abstract>

42

Kumble, Lithin, e Kiran Kumari Patil. "An Improved Stacked Auto Encoder based Data Compression Technique for WSNs". International Journal for Research in Applied Science and Engineering Technology 10, n. 5 (31 maggio 2022): 199–207. http://dx.doi.org/10.22214/ijraset.2022.41838.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

Abstract: Because of the limited energy available to sensor nodes in wireless sensor networks (WSNs), data compression is critical in these networks. The majority of the time, data communication results in energy consumption; however, by minimizing data transmission and reception, the lifetime of sensor nodes may usually be extended significantly. To compress sensor data, we present a new Improved Stacked RBM Auto-Encoder model, which is built of two layers: an encode layer and a decode layer, which is described in detail in this work. Data from sensors is compressed and decompressed in the encode layer; data from sensors is reconstructed and compressed in the decode layer. The encode layer and the decode layer are both made up of four conventional Restricted Boltzmann Machines that are used throughout the system (RBMs). We also present an energy optimization strategy that, by trimming the parameters of the model, can further minimize the energy consumption of the model storage and calculation. We evaluate the model's performance by comparing it to the data acquired by Intel Lab in the environment. Assuming that the model's compression ratio is 10, the average Percentage RMS Difference value is 9.84 percent, and the average temperature reconstruction error value is 0.312 degrees Celsius. It is possible to minimize the energy consumption of node communication in WSNs by 92 percent. When compared to the traditional method, the proposed model achieves higher compression efficiency and reconstruction accuracy while maintaining the same compression ratio as the old method. The results of our experiments demonstrate that the new neural network model can not only be applied to data compression for WSNs, but it also has high compression efficiency and an excellent transfer learning capability. Keywords: Data Compression; Stacked-Autocoder; transfer learning; energy, consumption optimization

43

Mannepalli, Kasiprasad, Suryabhan Pratap Singh, Chandra Sekhar Kolli, Sundeep Raj, Giridhar Reddy Bojja, B. R. Rajakumar e D. Binu. "Popularity Prediction Model With Context, Time and User Sentiment Information: An Optimization Assisted Deep Learning Technique". International Journal of Uncertainty, Fuzziness and Knowledge-Based Systems 31, n. 02 (aprile 2023): 283–302. http://dx.doi.org/10.1142/s0218488523500150.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

In social media, the data-sharing activities have turned out to be more pervasive; individuals and companies have comprehended the significance of promoting info by social media network. However, these individuals and companies face more challenges with the issue of “how to obtain the full benefit that the platforms provide”. Therefore, social media policies to improve the online promotion are turning out to be more significant. The popularization of social media contents are related to public attention and interest of users, thus the popularity fore cast of online contents has considered being the major task in social media analytic and it facilitates several appliances in diverse domain as well. This paper intends to introduce a popularity forecast approach that derives and combines the richest data of “text content encoder, user encoder, time series encoder, and user sentiment analysis”. The extracted features are then predicted via Long Short Term Memory (LSTM). Particularly, to enhance the prediction accuracy of the LSTM, the weights are fine-tuned via Self Adaptive Rain optimization (SA-RO).

44

Cruz, Angelo R. Dela, Ryan Rhay P. Vicerra, Argel A. Bandala e Elmer P. Dadios. "Dynamic Rate Allocation Algorithm Using Adaptive LMS End-to-End Distortion Estimation for Video Transmission over Error Prone Network". Journal of Advanced Computational Intelligence and Intelligent Informatics 20, n. 1 (19 gennaio 2016): 106–10. http://dx.doi.org/10.20965/jaciii.2016.p0106.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

Because of the inherent trade-off between source distortion and channel distortion in video transmission systems, joint optimization between bit-rate and distortion is still a challenging task. In this paper, we propose a method where the bit-rate allocation between source and channel encoder is controlled by the estimated end-to-end distortion at the encoder. The distortion estimation scheme is based on the adaptive forward linear predictor using least-mean square (LMS) algorithm. The forward predictor used the past values of actual end-to-end distortion to estimate the current distortion. The results show good estimate of end-to-end distortion and the proposed scheme improves video quality as compared to a standard rate control of H.264/AVC. The proposed scheme dynamically allocates the source encoder bits based on the estimated distortion.

45

George, Neema, e Anoop B. K. "Hypervolume Sen Task Scheduilng and Multi Objective Deep Auto Encoder based Resource Allocation in Cloud". International Journal on Recent and Innovation Trends in Computing and Communication 11, n. 4s (3 aprile 2023): 16–27. http://dx.doi.org/10.17762/ijritcc.v11i4s.6303.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

Cloud Computing (CC) environment has restructured the Information Age by empowering on demand dispensing of resources on a pay-per-use base. Resource Scheduling and allocation is an approach of ascertaining schedule on which tasks should be carried out. Owing to the heterogeneity nature of resources, scheduling of resources in CC environment is considered as an intricate task. Allocating best resource for a cloud request remains a complicated task and the issue of identifying the best resource – task pair according to user requirements is considered as an optimization issue. Therefore the main objective of the Cloud Server remains in scheduling the tasks and allocating the resources in an optimal manner. In this work an optimized task scheduled resource allocation model is designed to effectively address large numbers of task request arriving from cloud users, while maintaining enhanced Quality of Service (QoS). The cloud user task requests are mapped in an optimal manner to cloud resources. The optimization process is carried out using the proposed Multi-objective Auto-encoder Deep Neural Network-based (MA-DNN) method which is a combination of Sen’s Multi-objective functions and Auto-encoder Deep Neural Network model. First tasks scheduling is performed by applying Hypervolume-based Sen’s Multi-objective programming model. With this, multi-objective optimization (i.e., optimization of cost and time during the scheduling of tasks) is performed by means of Hypervolume-based Sen’s Multi-objective programming. Second, Auto-encoder Deep Neural Network-based Resource allocation is performed with the scheduled tasks that in turn allocate the resources by utilizing Jensen–Shannon divergence function. The Jensen–Shannon divergence function has the advantage of minimizing the energy consumption that only with higher divergence results, mapping is performed, therefore improving the energy consumption to a greater extent. Finally, mapping tasks with the corresponding resources using Kronecker Delta function improves the makespan significantly. To show the efficiency of Multi-objective Auto-encoder Deep Neural Network-based (MA-DNN) cloud time scheduling and optimization between tasks and resources in the CC environment, we also perform thorough experiments on the basis of realistic traces derived from Personal Cloud Datasets. The experimental results show that compared with RAA-PI-NSGAII and DRL, MA-DNN not only significantly accelerates the task scheduling efficiency, task scheduling time but also reduces the energy usage and makespan considerably.

46

SANDOVAL, C. "POWER CONSUMPTION OPTIMIZATION IN REED SOLOMON ENCODERS OVER FPGA". Latin American Applied Research - An international journal 44, n. 1 (31 gennaio 2014): 81–85. http://dx.doi.org/10.52292/j.laar.2014.422.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

This paper presents an analysis of the Reed Solomon encoder model and GF (2m) multiplier component, with the aim of optimizing the power consumption for reconfigurable hardware. The methods used consisted of concatenation and reassignment circuit signals in the VHDL description. This treatment allowed achieving a reduction in the consumption of hardware resources and optimizing power consumption in the multiplier of 7.89%, which results in a reduction of the dynamic power of a 42.42% in the coder design optimized. With this development, it provides a design method with good performance, which can be applied to other circuits.

47

Yu, Shengtao, Cheolkon Jung e Qiaozhou Lin. "HEVC encoder optimization for HDR video coding based on irregularity concealment effect". Signal Processing: Image Communication 64 (maggio 2018): 68–77. http://dx.doi.org/10.1016/j.image.2018.02.008.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

48

Trevithick, Alex, Matthew Chan, Michael Stengel, Eric Chan, Chao Liu, Zhiding Yu, Sameh Khamis, Manmohan Chandraker, Ravi Ramamoorthi e Koki Nagano. "Real-Time Radiance Fields for Single-Image Portrait View Synthesis". ACM Transactions on Graphics 42, n. 4 (26 luglio 2023): 1–15. http://dx.doi.org/10.1145/3592460.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

We present a one-shot method to infer and render a photorealistic 3D representation from a single unposed image (e.g., face portrait) in real-time. Given a single RGB input, our image encoder directly predicts a canonical triplane representation of a neural radiance field for 3D-aware novel view synthesis via volume rendering. Our method is fast (24 fps) on consumer hardware, and produces higher quality results than strong GAN-inversion baselines that require test-time optimization. To train our triplane encoder pipeline, we use only synthetic data, showing how to distill the knowledge from a pretrained 3D GAN into a feedforward encoder. Technical contributions include a Vision Transformer-based triplane encoder, a camera data augmentation strategy, and a well-designed loss function for synthetic data training. We benchmark against the state-of-the-art methods, demonstrating significant improvements in robustness and image quality in challenging real-world settings. We showcase our results on portraits of faces (FFHQ) and cats (AFHQ), but our algorithm can also be applied in the future to other categories with a 3D-aware image generator.

49

Yan, Xiaoan, Yadong Xu, Daoming She e Wan Zhang. "Reliable Fault Diagnosis of Bearings Using an Optimized Stacked Variational Denoising Auto-Encoder". Entropy 24, n. 1 (24 dicembre 2021): 36. http://dx.doi.org/10.3390/e24010036.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

Variational auto-encoders (VAE) have recently been successfully applied in the intelligent fault diagnosis of rolling bearings due to its self-learning ability and robustness. However, the hyper-parameters of VAEs depend, to a significant extent, on artificial settings, which is regarded as a common and key problem in existing deep learning models. Additionally, its anti-noise capability may face a decline when VAE is used to analyze bearing vibration data under loud environmental noise. Therefore, in order to improve the anti-noise performance of the VAE model and adaptively select its parameters, this paper proposes an optimized stacked variational denoising autoencoder (OSVDAE) for the reliable fault diagnosis of bearings. Within the proposed method, a robust network, named variational denoising auto-encoder (VDAE), is, first, designed by integrating VAE and a denoising auto-encoder (DAE). Subsequently, a stacked variational denoising auto-encoder (SVDAE) architecture is constructed to extract the robust and discriminative latent fault features via stacking VDAE networks layer on layer, wherein the important parameters of the SVDAE model are automatically determined by employing a novel meta-heuristic intelligent optimizer known as the seagull optimization algorithm (SOA). Finally, the extracted latent features are imported into a softmax classifier to obtain the results of fault recognition in rolling bearings. Experiments are conducted to validate the effectiveness of the proposed method. The results of analysis indicate that the proposed method not only can achieve a high identification accuracy for different bearing health conditions, but also outperforms some representative deep learning methods.

50

Yang, Dong, Jingyuan Wang e Xi Yang. "3D Point Cloud Shape Generation with Collaborative Learning of Generative Adversarial Network and Auto-Encoder". Remote Sensing 16, n. 10 (16 maggio 2024): 1772. http://dx.doi.org/10.3390/rs16101772.

Testo completo

Gli stili APA, Harvard, Vancouver, ISO e altri

Abstract (sommario):

A point cloud is a simple and concise 3D representation, but point cloud generation is a long-term challenging task in 3D vision. However, most existing methods only focus on their effectiveness of generation and auto-encoding separately. Furthermore, both generative adversarial networks (GANs) and auto-encoders (AEs) are the most popular generative models. But there is a lack of related research that investigates the implicit connections between them in the field of point cloud generation. Thus, we propose a new bidirectional network (BI-Net) trained with collaborative learning, introducing more priors through the alternate parameter optimizations of a GAN and AE combination, which is different from the way of combining them at the network structure and loss function level. Specifically, BI-Net acts as a GAN and AE in different data processing directions, where their network structures can be reused. If optimizing only the GAN without the AE, there is no direct constraint of ground truth on the generator’s parameter optimization. This unique approach enables better network optimization and leads to superior generation results. Moreover, we propose a nearest neighbor mutual exclusion (NNME) loss to further homogenize the spatial distribution of generated points during the reverse direction. Extensive experiments were conducted, and the results show that the BI-Net produces competitive and high-quality results on reasonable structure and uniform distributions compared to existing state-of-the-art methods. We believe that our network structure (BI-Net) with collaborative learning could provide a new promising method for future point cloud generation tasks.

Articoli di riviste sul tema "Encoder optimization"

Cita una fonte nei formati APA, MLA, Chicago, Harvard e in molti altri stili