arXiv subjects

Yongjun Xu

Publications and source records attributed to Yongjun Xu.

At least 109 records · Page 6Linked to original sources

arXiv2022

Knowledge Distillation Using Hierarchical Self-Supervision Augmented Distribution

Knowledge distillation (KD) is an effective framework that aims to transfer meaningful information from a large teacher to a smaller student. Generally, KD often involves how to define and transfer knowledge. Previous KD methods often focus on mining various forms of knowledge, for example, feature maps and refined information. However, the knowledge is derived from the primary supervised task and thus is highly task-specific. Motivated by the recent success of self-supervised representation learning, we propose an auxiliary self-supervision augmented task to guide networks to learn more meaningful features. Therefore, we can derive soft self-supervision augmented distributions as richer dark knowledge from this task for KD. Unlike previous knowledge, this distribution encodes joint knowledge from supervised and self-supervised feature learning. Beyond knowledge exploration, we propose to append several auxiliary branches at various hidden layers, to fully take advantage of hierarchical feature maps. Each auxiliary branch is guided to learn self-supervision augmented task and distill this distribution from teacher to student. Overall, we call our KD method as Hierarchical Self-Supervision Augmented Knowledge Distillation (HSSAKD). Experiments on standard image classification show that both offline and online HSSAKD achieves state-of-the-art performance in the field of KD. Further transfer experiments on object detection further verify that HSSAKD can guide the network to learn better features. The code is available at https://github.com/winycg/HSAKD.

cs.CV↗

arXiv2022

Hierarchical Self-supervised Augmented Knowledge Distillation

Knowledge distillation often involves how to define and transfer knowledge from teacher to student effectively. Although recent self-supervised contrastive knowledge achieves the best performance, forcing the network to learn such knowledge may damage the representation learning of the original class recognition task. We therefore adopt an alternative self-supervised augmented task to guide the network to learn the joint distribution of the original recognition task and self-supervised auxiliary task. It is demonstrated as a richer knowledge to improve the representation power without losing the normal classification capability. Moreover, it is incomplete that previous methods only transfer the probabilistic knowledge between the final layers. We propose to append several auxiliary classifiers to hierarchical intermediate feature maps to generate diverse self-supervised knowledge and perform the one-to-one transfer to teach the student network thoroughly. Our method significantly surpasses the previous SOTA SSKD with an average improvement of 2.56\% on CIFAR-100 and an improvement of 0.77\% on ImageNet across widely used network pairs. Codes are available at https://github.com/winycg/HSAKD.

cs.CV↗

arXiv2022

Localizing Semantic Patches for Accelerating Image Classification

Existing works often focus on reducing the architecture redundancy for accelerating image classification but ignore the spatial redundancy of the input image. This paper proposes an efficient image classification pipeline to solve this problem. We first pinpoint task-aware regions over the input image by a lightweight patch proposal network called AnchorNet. We then feed these localized semantic patches with much smaller spatial redundancy into a general classification network. Unlike the popular design of deep CNN, we aim to carefully design the Receptive Field of AnchorNet without intermediate convolutional paddings. This ensures the exact mapping from a high-level spatial location to the specific input image patch. The contribution of each patch is interpretable. Moreover, AnchorNet is compatible with any downstream architecture. Experimental results on ImageNet show that our method outperforms SOTA dynamic inference methods with fewer inference costs. Our code is available at https://github.com/winycg/AnchorNet.

cs.CV↗

arXiv2022

PFGDF: Pruning Filter via Gaussian Distribution Feature for Deep Neural Networks Acceleration

Deep learning has achieved impressive results in many areas, but the deployment of edge intelligent devices is still very slow. To solve this problem, we propose a novel compression and acceleration method based on data distribution characteristics for deep neural networks, namely Pruning Filter via Gaussian Distribution Feature (PFGDF). Compared with previous advanced pruning methods, PFGDF compresses the model by filters with insignificance in distribution, regardless of the contribution and sensitivity information of the convolution filter. PFGDF is significantly different from weight sparsification pruning because it does not require the special accelerated library to process the sparse weight matrix and introduces no more extra parameters. The pruning process of PFGDF is automated. Furthermore, the model compressed by PFGDF can restore the same performance as the uncompressed model. We evaluate PFGDF through extensive experiments, on CIFAR-10, PFGDF compresses the convolution filter on VGG-16 by 66.62% with more than 90% parameter reduced, while the inference time is accelerated by 83.73% on Huawei MATE 10.

cs.CV↗

arXiv2022

Cross-Image Relational Knowledge Distillation for Semantic Segmentation

Current Knowledge Distillation (KD) methods for semantic segmentation often guide the student to mimic the teacher's structured information generated from individual data samples. However, they ignore the global semantic relations among pixels across various images that are valuable for KD. This paper proposes a novel Cross-Image Relational KD (CIRKD), which focuses on transferring structured pixel-to-pixel and pixel-to-region relations among the whole images. The motivation is that a good teacher network could construct a well-structured feature space in terms of global pixel dependencies. CIRKD makes the student mimic better structured semantic relations from the teacher, thus improving the segmentation performance. Experimental results over Cityscapes, CamVid and Pascal VOC datasets demonstrate the effectiveness of our proposed approach against state-of-the-art distillation methods. The code is available at https://github.com/winycg/CIRKD.

cs.CV↗

arXiv2022

Lifelong Generative Learning via Knowledge Reconstruction

Generative models often incur the catastrophic forgetting problem when they are used to sequentially learning multiple tasks, i.e., lifelong generative learning. Although there are some endeavors to tackle this problem, they suffer from high time-consumptions or error accumulation. In this work, we develop an efficient and effective lifelong generative model based on variational autoencoder (VAE). Unlike the generative adversarial network, VAE enjoys high efficiency in the training process, providing natural benefits with few resources. We deduce a lifelong generative model by expending the intrinsic reconstruction character of VAE to the historical knowledge retention. Further, we devise a feedback strategy about the reconstructed data to alleviate the error accumulation. Experiments on the lifelong generating tasks of MNIST, FashionMNIST, and SVHN verified the efficacy of our approach, where the results were comparable to SOTA.

cs.LG↗

arXiv2022

Mutual Contrastive Learning for Visual Representation Learning

We present a collaborative learning method called Mutual Contrastive Learning (MCL) for general visual representation learning. The core idea of MCL is to perform mutual interaction and transfer of contrastive distributions among a cohort of networks. A crucial component of MCL is Interactive Contrastive Learning (ICL). Compared with vanilla contrastive learning, ICL can aggregate cross-network embedding information and maximize the lower bound to the mutual information between two networks. This enables each network to learn extra contrastive knowledge from others, leading to better feature representations for visual recognition tasks. We emphasize that the resulting MCL is conceptually simple yet empirically powerful. It is a generic framework that can be applied to both supervised and self-supervised representation learning. Experimental results on image classification and transfer learning to object detection show that MCL can lead to consistent performance gains, demonstrating that MCL can guide the network to generate better feature representations. Code is available at https://github.com/winycg/MCL.

cs.CV↗

arXiv2021

Towards Robust Knowledge Graph Embedding via Multi-task Reinforcement Learning

Nowadays, Knowledge graphs (KGs) have been playing a pivotal role in AI-related applications. Despite the large sizes, existing KGs are far from complete and comprehensive. In order to continuously enrich KGs, automatic knowledge construction and update mechanisms are usually utilized, which inevitably bring in plenty of noise. However, most existing knowledge graph embedding (KGE) methods assume that all the triple facts in KGs are correct, and project both entities and relations into a low-dimensional space without considering noise and knowledge conflicts. This will lead to low-quality and unreliable representations of KGs. To this end, in this paper, we propose a general multi-task reinforcement learning framework, which can greatly alleviate the noisy data problem. In our framework, we exploit reinforcement learning for choosing high-quality knowledge triples while filtering out the noisy ones. Also, in order to take full advantage of the correlations among semantically similar relations, the triple selection processes of similar relations are trained in a collective way with multi-task learning. Moreover, we extend popular KGE models TransE, DistMult, ConvE and RotatE with the proposed framework. Finally, the experimental validation shows that our approach is able to enhance existing KGE models and can provide more robust representations of KGs in noisy scenarios.

cs.CL↗

arXiv2021

Time-Varying Channel Prediction for RIS-Assisted MU-MISO Networks via Deep Learning

To mitigate the effects of shadow fading and obstacle blocking, reconfigurable intelligent surface (RIS) has become a promising technology to improve the signal transmission quality of wireless communications by controlling the reconfigurable passive elements with less hardware cost and lower power consumption. However, accurate, low-latency and low-pilot-overhead channel state information (CSI) acquisition remains a considerable challenge in RIS-assisted systems due to the large number of RIS passive elements. In this paper, we propose a three-stage joint channel decomposition and prediction framework to require CSI. The proposed framework exploits the two-timescale property that the base station (BS)-RIS channel is quasi-static and the RIS-user equipment (UE) channel is fast time-varying. Specifically, in the first stage, we use the full-duplex technique to estimate the channel between a BS's specific antenna and the RIS, addressing the critical scaling ambiguity problem in the channel decomposition. We then design a novel deep neural network, namely, the sparse-connected long short-term memory (SCLSTM), and propose a SCLSTM-based algorithm in the second and third stages, respectively. The algorithm can simultaneously decompose the BS-RIS channel and RIS-UE channel from the cascaded channel and capture the temporal relationship of the RIS-UE channel for prediction. Simulation results show that our proposed framework has lower pilot overhead than the traditional channel estimation algorithms, and the proposed SCLSTM-based algorithm can also achieve more accurate CSI acquisition robustly and effectively.

cs.IT↗

arXiv2021

Convolutional Autoencoder-Based Phase Shift Feedback Compression for Intelligent Reflecting Surface-Assisted Wireless Systems

In recent years, intelligent reflecting surface (IRS) has emerged as a promising technology for 6G due to its potential/ability to significantly enhance energy- and spectrum-efficiency. To this end, it is crucial to adjust the phases of reflecting elements of the IRS, and most of the research works focus on how to optimize/quantize the phase for different optimization objectives. In particular, the quantized phase shift (QPS) is assumed to be available at the IRS, which, however, does not always hold and should be fed back to the IRS in practice. Unfortunately, the feedback channel is generally bandwidth-limited, which cannot support a huge amount of feedback overhead of the QPS particularly for a large number of reflecting elements and/or the quantization level of each reflecting element. In order to break this bottleneck, in this letter, we propose a convolutional autoencoder-based scheme, in which the QPS is compressed on the receiver side and reconstructed on the IRS side. In order to solve the problems of mismatched distribution and vanishing gradient, we remove the batch normalization (BN) layers and introduce a denosing module. By doing so, it is possible to achieve a high compression ratio with a reliable reconstruction accuracy in the bandwidth-limited feedback channel, and it is also possible to accommodate existing works assuming available QPS at the IRS. Simulation results confirm the high reconstruction accuracy of the feedback/compressed QPS through a feedback channel, and show that the proposed scheme can significantly outperform the existing compression algorithms.

eess.SP↗

arXiv2021

A Channel-Aware Routing Protocol With Nearest Neighbor Regression For Underwater Sensor Networks

The underwater acoustic channel is one of the most challenging communication channels. Due to periodical tidal and daily climatic variation, underwater noise is periodically fluctuating, which result in the periodical changing of acoustic channel quality in long-term. Also, time-variant channel quality leads to routing failure. Routing protocols with acoustic channel estimation, namely underwater channel-aware routing protocols are recently proposed to maintain the routing performance. However, channel estimation algorithms for these routing protocols are mostly linear and rarely consider periodicity of acoustic channels. In this paper, we introduce acoustic channel estimation based on nearest neighbor regression for underwater acoustic networks. We extend nearest neighbor regression for SNR (Signal-to-Noise Ratio) time series prediction, providing an outstanding prediction accuracy for intricately periodical and fluctuating received SNR time series. Moreover, we propose a quick search algorithm and use statistical storage compression to optimize the time and space complexity of the algorithm. In contrast with linear methods, this algorithm significantly improves channel prediction accuracy (over three times at most) on both simulation and sea trial data sets. With this channel estimation method, we then propose a Depth-Based Channel-Aware Routing protocol (DBCAR). Taking advantage of depth-greedy forwarding and channel-aware reliable communication, DBCAR has an outstanding network performance on packet delivery ratio, average energy consumption and average transmission delay which is validated through extensive simulations.

cs.NI↗

arXiv2021

A Distributed SGD Algorithm with Global Sketching for Deep Learning Training Acceleration

Distributed training is an effective way to accelerate the training process of large-scale deep learning models. However, the parameter exchange and synchronization of distributed stochastic gradient descent introduce a large amount of communication overhead. Gradient compression is an effective method to reduce communication overhead. In synchronization SGD compression methods, many Top-k sparsification based gradient compression methods have been proposed to reduce the communication. However, the centralized method based on the parameter servers has the single point of failure problem and limited scalability, while the decentralized method with global parameter exchanging may reduce the convergence rate of training. In contrast with Top-$k$ based methods, we proposed a gradient compression method with globe gradient vector sketching, which uses the Count-Sketch structure to store the gradients to reduce the loss of the accuracy in the training process, named global-sketching SGD (gs-SGD). The gs-SGD has better convergence efficiency on deep learning models and a communication complexity of O($\log d*\log P$), where $d$ is the number of model parameters and P is the number of workers. We conducted experiments on GPU clusters to verify that our method has better convergence efficiency than global Top-$k$ and Sketching-based methods. In addition, gs-SGD achieves 1.3-3.1x higher throughput compared with gTop-$k$, and 1.1-1.2x higher throughput compared with original Sketched-SGD.

cs.DC↗

arXiv2021

Channel Pruning via Multi-Criteria based on Weight Dependency

Channel pruning has demonstrated its effectiveness in compressing ConvNets. In many related arts, the importance of an output feature map is only determined by its associated filter. However, these methods ignore a small part of weights in the next layer which disappears as the feature map is removed. They ignore the phenomenon of weight dependency. Besides, many pruning methods use only one criterion for evaluation and find a sweet spot of pruning structure and accuracy in a trial-and-error fashion, which can be time-consuming. In this paper, we proposed a channel pruning algorithm via multi-criteria based on weight dependency, CPMC, which can compress a pre-trained model directly. CPMC defines channel importance in three aspects, including its associated weight value, computational cost, and parameter quantity. According to the phenomenon of weight dependency, CPMC gets channel importance by assessing its associated filter and the corresponding partial weights in the next layer. Then CPMC uses global normalization to achieve cross-layer comparison. Finally, CPMC removes less important channels by global ranking. CPMC can compress various CNN models, including VGGNet, ResNet, and DenseNet on various image classification datasets. Extensive experiments have shown CPMC outperforms the others significantly.

cs.CV↗

arXiv2021

Multi-view Contrastive Learning for Online Knowledge Distillation

Previous Online Knowledge Distillation (OKD) often carries out mutually exchanging probability distributions, but neglects the useful representational knowledge. We therefore propose Multi-view Contrastive Learning (MCL) for OKD to implicitly capture correlations of feature embeddings encoded by multiple peer networks, which provide various views for understanding the input data instances. Benefiting from MCL, we can learn a more discriminative representation space for classification than previous OKD methods. Experimental results on image classification demonstrate that our MCL-OKD outperforms other state-of-the-art OKD methods by large margins without sacrificing additional inference cost. Codes are available at https://github.com/winycg/MCL-OKD.

cs.CV↗

arXiv2020

GHFP: Gradually Hard Filter Pruning

Filter pruning is widely used to reduce the computation of deep learning, enabling the deployment of Deep Neural Networks (DNNs) in resource-limited devices. Conventional Hard Filter Pruning (HFP) method zeroizes pruned filters and stops updating them, thus reducing the search space of the model. On the contrary, Soft Filter Pruning (SFP) simply zeroizes pruned filters, keeping updating them in the following training epochs, thus maintaining the capacity of the network. However, SFP, together with its variants, converges much slower than HFP due to its larger search space. Our question is whether SFP-based methods and HFP can be combined to achieve better performance and speed up convergence. Firstly, we generalize SFP-based methods and HFP to analyze their characteristics. Then we propose a Gradually Hard Filter Pruning (GHFP) method to smoothly switch from SFP-based methods to HFP during training and pruning, thus maintaining a large search space at first, gradually reducing the capacity of the model to ensure a moderate convergence speed. Experimental results on CIFAR-10/100 show that our method achieves the state-of-the-art performance.

cs.CV↗

arXiv2020

Softer Pruning, Incremental Regularization

Network pruning is widely used to compress Deep Neural Networks (DNNs). The Soft Filter Pruning (SFP) method zeroizes the pruned filters during training while updating them in the next training epoch. Thus the trained information of the pruned filters is completely dropped. To utilize the trained pruned filters, we proposed a SofteR Filter Pruning (SRFP) method and its variant, Asymptotic SofteR Filter Pruning (ASRFP), simply decaying the pruned weights with a monotonic decreasing parameter. Our methods perform well across various networks, datasets and pruning rates, also transferable to weight pruning. On ILSVRC-2012, ASRFP prunes 40% of the parameters on ResNet-34 with 1.63% top-1 and 0.68% top-5 accuracy improvement. In theory, SRFP and ASRFP are an incremental regularization of the pruned filters. Besides, We note that SRFP and ASRFP pursue better results while slowing down the speed of convergence.

cs.CV↗

arXiv2020

Localizing Interpretable Multi-scale informative Patches Derived from Media Classification Task

Deep convolutional neural networks (CNN) always depend on wider receptive field (RF) and more complex non-linearity to achieve state-of-the-art performance, while suffering the increased difficult to interpret how relevant patches contribute the final prediction. In this paper, we construct an interpretable AnchorNet equipped with our carefully designed RFs and linearly spatial aggregation to provide patch-wise interpretability of the input media meanwhile localizing multi-scale informative patches only supervised on media-level labels without any extra bounding box annotations. Visualization of localized informative image and text patches show the superior multi-scale localization capability of AnchorNet. We further use localized patches for downstream classification tasks across widely applied networks. Experimental results demonstrate that replacing the original inputs with their patches for classification can get a clear inference acceleration with only tiny performance degradation, which proves that localized patches can indeed retain the most semantics and evidences of the original inputs.

cs.CV↗

arXiv2020

Towards Green Mobile Edge Computing Offloading Systems with Security Enhancement

Mobile edge computing (MEC) is an emerging communication scheme that aims at reducing latency. In this paper, we investigate a green MEC system under the existence of an eavesdropper. We use computation efficiency, which is defined as the total number of bits computed divided by the total energy consumption, as our main metric. To alleviate stringent latency requirement, a joint secure offloading and computation model is proposed. Additionally, we formulate an optimization problem for maximizing computation efficiency, under several practical constraints. The non-convex problem is tackled by successive convex approximation and an iterative algorithm. Simulation results have verified the superiority of our proposed scheme, as well as the effectiveness of our problem solution.

cs.NI↗