The Experts below are selected from a list of 14544 Experts worldwide ranked by ideXlab platform

Kai Chen - One of the best experts on this subject based on the ideXlab platform.

  • Convolutional Regression for Visual Tracking
    IEEE Transactions on Image Processing, 2018
    Co-Authors: Kai Chen
    Abstract:

    Recently, discriminatively learned correlation filters (DCF) has attracted much attention in visual object tracking community. The success of DCF is potentially attributed to the fact that a large number of samples are utilized to train the ridge regression model and predict the location of an object. To solve the regression problem in an efficient way, these samples are all generated by circularly shifting from a searching patch. However, these synthetic samples also induce some negative effects that weaken the robustness of DCF-based trackers. In this paper, we propose a new approach to learn the regression model for visual tracking with single Convolutional Layer. Instead of learning the linear regression model in a closed form, we try to solve the regression problem by optimizing a one-channel-output Convolution Layer with gradient descent (GD). In particular, the kernel size of the Convolution Layer is set to the size of the object. Contrary to DCF, it is possible to incorporate all “real” samples clipped from the whole image. A critical issue of the GD approach is that most of the Convolutional samples are negative and the contribution of positive samples will be suppressed. To address this problem, we propose a novel objective function to eliminate easy negatives and enhance positives. We perform extensive experiments on four widely used datasets: OTB-100, OTB-50, TempleColor, and VOT-2016. The results show that the proposed algorithm achieves outstanding performance and outperforms most of the existing DCF-based algorithms.

  • Convolutional Neural Networks for Page Segmentation of Historical Document Images
    2017 14th IAPR International Conference on Document Analysis and Recognition (ICDAR), 2017
    Co-Authors: Kai Chen, Mathias Seuret, Jean Hennebert, Rolf Ingold
    Abstract:

    This paper presents a page segmentation method for handwritten historical document images based on a Convolutional Neural Network (CNN). We consider page segmentation as a pixel labeling problem, i.e., each pixel is classified as one of the predefined classes. Traditional methods in this area rely on hand-crafted features carefully tuned considering prior knowledge. In contrast, we propose to learn features from raw image pixels using a CNN. While many researchers focus on developing deep CNN architectures to solve different problems, we train a simple CNN with only one Convolution Layer. We show that the simple architecture achieves competitive results against other deep architectures on different public datasets. Experiments also demonstrate the effectiveness and superiority of the proposed method compared to previous methods.

Ha-jin Yu - One of the best experts on this subject based on the ideXlab platform.

  • A Complete End-to-End Speaker Verification System Using Deep Neural Networks: From Raw Signals to Verification Result
    2018 IEEE International Conference on Acoustics Speech and Signal Processing (ICASSP), 2018
    Co-Authors: Jee- Weon Jung, Il-ho Yang, Hye-jin Shim, Ha-jin Yu
    Abstract:

    End-to-end systems using deep neural networks have been widely studied in the field of speaker verification. Raw audio signal processing has also been widely studied in the fields of automatic music tagging and speech recognition. However, as far as we know, end-to-end systems using raw audio signals have not been explored in speaker verification. In this paper, a complete end-to-end speaker verification system is proposed, which inputs raw audio signals and outputs the verification results. A pre-processing Layer and the embedded speaker feature extraction models were mainly investigated. The proposed pre-emphasis Layer was combined with a strided Convolution Layer for pre-processing at the first two hidden Layers. In addition, speaker feature extraction models using ConvolutionalLayer and long short-term memory are proposed to be embedded in the proposed end-to-end system.

Masatoshi Okutomi - One of the best experts on this subject based on the ideXlab platform.

  • Coupled Convolution Layer for Convolutional neural network.
    Neural Networks, 2018
    Co-Authors: Kazutaka Uchida, Masayuki Tanaka, Masatoshi Okutomi
    Abstract:

    Abstract We propose a coupled Convolution Layer comprising multiple parallel Convolutions with mutually constrained filters. Inspired by biological human vision mechanism, we constrain the Convolution filters such that one set of filter weights should be geometrically rotated, mirrored, or be the negative of the other. Our analysis suggests that the coupled Convolution Layer is more effective for lower Layer where feature maps preserve geometric properties. Experimental comparisons demonstrate that the proposed coupled Convolution Layer performs slightly better than the original Layer while decreasing the number of parameters. We evaluate its effect compared to non-constrained Convolution Layer using the CIFAR-10, CIFAR-100, and PlanktonSet 1.0 datasets.

  • ICPR - Coupled Convolution Layer for Convolutional neural network
    2016 23rd International Conference on Pattern Recognition (ICPR), 2016
    Co-Authors: Kazutaka Uchida, Masayuki Tanaka, Masatoshi Okutomi
    Abstract:

    We introduce a coupled Convolution Layer comprising two parallel Convolutions with mutually constrained weights. Inspired by the human retina mechanism, we constrain the Convolution weights such that one set of weights should be the negative of the other to mimic responses of on-center and off-center retinal ganglion cells. Our analysis shows that the retina-like Convolution Layer, a special case of the coupled Convolution Layer, can be realized by a normal Convolutional Layer with a pair of activation functions designated as Biased ON/OFF ReLU. Experimental comparisons demonstrate that the proposed coupled Convolution Layer performs better without increasing the number of parameters, which reveals two important facts. First, the separation of the positive and negative part into different channels plays an important role. Secondly, constraining weights across Convolutions can produce better performance than training weights freely. We evaluate its effect by comparison with ReLU, LReLU, and PReLU using the CIFAR-10, CIFAR-100, and PlanktonSet 1.0 datasets.

  • Coupled Convolution Layer for Convolutional neural network
    2016 23rd International Conference on Pattern Recognition (ICPR), 2016
    Co-Authors: Kazutaka Uchida, Masayuki Tanaka, Masatoshi Okutomi
    Abstract:

    We introduce a coupled Convolution Layer comprising two parallel Convolutions with mutually constrained weights. Inspired by the human retina mechanism, we constrain the Convolution weights such that one set of weights should be the negative of the other to mimic responses of on-center and off-center retinal ganglion cells. Our analysis shows that the retina-like Convolution Layer, a special case of the coupled Convolution Layer, can be realized by a normal Convolutional Layer with a pair of activation functions designated as Biased ON/OFF ReLU. Experimental comparisons demonstrate that the proposed coupled Convolution Layer performs better without increasing the number of parameters, which reveals two important facts. First, the separation of the positive and negative part into different channels plays an important role. Secondly, constraining weights across Convolutions can produce better performance than training weights freely. We evaluate its effect by comparison with ReLU, LReLU, and PReLU using the CIFAR-10, CIFAR-100, and PlanktonSet 1.0 datasets.

Zhi Yang - One of the best experts on this subject based on the ideXlab platform.

  • SeerNet: Predicting Convolutional Neural Network Feature-Map Sparsity Through Low-Bit Quantization
    2019 IEEE CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019
    Co-Authors: Wencong Xiao, Lintao Zhang, Chen Zhang, Zhi Yang
    Abstract:

    In this paper we present a novel and general method to accelerate Convolutional neural network (CNN) inference by taking advantage of feature map sparsity. We experimentally demonstrate that a highly quantized version of the original network is sufficient in predicting the output sparsity accurately, and verify that leveraging such sparsity in inference incurs negligible accuracy drop compared with the original network. To accelerate inference, for each Convolution Layer our approach first obtains a binary sparsity mask of the output feature maps by running inference on a quantized version of the original network Layer, and then conducts a full-precision sparse Convolution to find out the precise values of the non-zero outputs. Compared with existing work, our approach avoids the overhead of training additional auxiliary networks, while is still applicable to general CNN networks without being limited to certain application domains.

  • CVPR - SeerNet: Predicting Convolutional Neural Network Feature-Map Sparsity Through Low-Bit Quantization
    2019 IEEE CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2019
    Co-Authors: Lingxiao Ma, Wencong Xiao, Lintao Zhang, Chen Zhang, Zhi Yang
    Abstract:

    In this paper we present a novel and general method to accelerate Convolutional neural network (CNN) inference by taking advantage of feature map sparsity. We experimentally demonstrate that a highly quantized version of the original network is sufficient in predicting the output sparsity accurately, and verify that leveraging such sparsity in inference incurs negligible accuracy drop compared with the original network. To accelerate inference, for each Convolution Layer our approach first obtains a binary sparsity mask of the output feature maps by running inference on a quantized version of the original network Layer, and then conducts a full-precision sparse Convolution to find out the precise values of the non-zero outputs. Compared with existing work, our approach avoids the overhead of training additional auxiliary networks, while is still applicable to general CNN networks without being limited to certain application domains.

A. Dalvadi Jigar - One of the best experts on this subject based on the ideXlab platform.

  • Swift single image super resolution using deep Convolution neural network
    2016 International Conference on Communication and Electronics Systems (ICCES), 2016
    Co-Authors: A. Dalvadi Jigar
    Abstract:

    Now a days, Image Super resolution is one of the most important and challenging issue in the image processing area. Aim of Super resolution is to generate high-resolution image from single or multiple low resolution of the same scene or image. With single low resolution image it's very challenging to produce high-resolution image because a single low-resolution image contain the less information. Due to the ability of preserving edges, kind of method called TV (Total Variation)-based method was proposed as regularization function for some inverse problems. Due to ill-posed nature of problem, existing super resolution method which based on combine total variation regularization term including the Non Local Total Variation (NLTV) and Steering Kernel Regularization Total Variation (SKRTV) which takes the more execution time due to the non-local weight calculation. We propose the Example based Convolution neural network which consists of three Layers namely Convolution Layer, max-pooling Layer and reconstruction Layer. Using Convolution neural network approach we achieved to reduce the execution time as well as increased the PSNR ratio.