arXiv subjects

Yonghong Tian

Publications and source records attributed to Yonghong Tian.

At least 217 records · Page 12Linked to original sources

arXiv2020

Self-Guided Adaptation: Progressive Representation Alignment for Domain Adaptive Object Detection

Unsupervised domain adaptation (UDA) has achieved unprecedented success in improving the cross-domain robustness of object detection models. However, existing UDA methods largely ignore the instantaneous data distribution during model learning, which could deteriorate the feature representation given large domain shift. In this work, we propose a Self-Guided Adaptation (SGA) model, target at aligning feature representation and transferring object detection models across domains while considering the instantaneous alignment difficulty. The core of SGA is to calculate "hardness" factors for sample pairs indicating domain distance in a kernel space. With the hardness factor, the proposed SGA adaptively indicates the importance of samples and assigns them different constrains. Indicated by hardness factors, Self-Guided Progressive Sampling (SPS) is implemented in an "easy-to-hard" way during model adaptation. Using multi-stage convolutional features, SGA is further aggregated to fully align hierarchical representations of detection models. Extensive experiments on commonly used benchmarks show that SGA improves the state-of-the-art methods with significant margins, while demonstrating the effectiveness on large domain shift.

cs.CV↗

arXiv2020

HRank: Filter Pruning using High-Rank Feature Map

Neural network pruning offers a promising prospect to facilitate deploying deep neural networks on resource-limited devices. However, existing methods are still challenged by the training inefficiency and labor cost in pruning designs, due to missing theoretical guidance of non-salient network components. In this paper, we propose a novel filter pruning method by exploring the High Rank of feature maps (HRank). Our HRank is inspired by the discovery that the average rank of multiple feature maps generated by a single filter is always the same, regardless of the number of image batches CNNs receive. Based on HRank, we develop a method that is mathematically formulated to prune filters with low-rank feature maps. The principle behind our pruning is that low-rank feature maps contain less information, and thus pruned results can be easily reproduced. Besides, we experimentally show that weights with high-rank feature maps contain more important information, such that even when a portion is not updated, very little damage would be done to the model performance. Without introducing any additional constraints, HRank leads to significant improvements over the state-of-the-arts in terms of FLOPs and parameters reduction, with similar accuracies. For example, with ResNet-110, we achieve a 58.2%-FLOPs reduction by removing 59.2% of the parameters, with only a small loss of 0.14% in top-1 accuracy on CIFAR-10. With Res-50, we achieve a 43.8%-FLOPs reduction by removing 36.7% of the parameters, with only a loss of 1.17% in the top-1 accuracy on ImageNet. The codes can be available at https://github.com/lmbxmu/HRank.

cs.CV↗

arXiv2020

Probabilistic Inference of Binary Markov Random Fields in Spiking Neural Networks through Mean-field Approximation

Recent studies have suggested that the cognitive process of the human brain is realized as probabilistic inference and can be further modeled by probabilistic graphical models like Markov random fields. Nevertheless, it remains unclear how probabilistic inference can be implemented by a network of spiking neurons in the brain. Previous studies have tried to relate the inference equation of binary Markov random fields to the dynamic equation of spiking neural networks through belief propagation algorithm and reparameterization, but they are valid only for Markov random fields with limited network structure. In this paper, we propose a spiking neural network model that can implement inference of arbitrary binary Markov random fields. Specifically, we design a spiking recurrent neural network and prove that its neuronal dynamics are mathematically equivalent to the inference process of Markov random fields by adopting mean-field theory. Furthermore, our mean-field approach unifies previous works. Theoretical analysis and experimental results, together with the application to image denoising, demonstrate that our proposed spiking neural network can get comparable results to that of mean-field inference.

q-bio.NC↗

arXiv2020

Neural System Identification with Spike-triggered Non-negative Matrix Factorization

Neuronal circuits formed in the brain are complex with intricate connection patterns. Such complexity is also observed in the retina as a relatively simple neuronal circuit. A retinal ganglion cell receives excitatory inputs from neurons in previous layers as driving forces to fire spikes. Analytical methods are required that can decipher these components in a systematic manner. Recently a method termed spike-triggered non-negative matrix factorization (STNMF) has been proposed for this purpose. In this study, we extend the scope of the STNMF method. By using the retinal ganglion cell as a model system, we show that STNMF can detect various computational properties of upstream bipolar cells, including spatial receptive field, temporal filter, and transfer nonlinearity. In addition, we recover synaptic connection strengths from the weight matrix of STNMF. Furthermore, we show that STNMF can separate spikes of a ganglion cell into a few subsets of spikes where each subset is contributed by one presynaptic bipolar cell. Taken together, these results corroborate that STNMF is a useful method for deciphering the structure of neuronal circuits.

q-bio.NC↗

arXiv2020

Revealing Fine Structures of the Retinal Receptive Field by Deep Learning Networks

Deep convolutional neural networks (CNNs) have demonstrated impressive performance on many visual tasks. Recently, they became useful models for the visual system in neuroscience. However, it is still not clear what are learned by CNNs in terms of neuronal circuits. When a deep CNN with many layers is used for the visual system, it is not easy to compare the structure components of CNNs with possible neuroscience underpinnings due to highly complex circuits from the retina to higher visual cortex. Here we address this issue by focusing on single retinal ganglion cells with biophysical models and recording data from animals. By training CNNs with white noise images to predict neuronal responses, we found that fine structures of the retinal receptive field can be revealed. Specifically, convolutional filters learned are resembling biological components of the retinal circuit. This suggests that a CNN learning from one single retinal cell reveals a minimal neural network carried out in this cell. Furthermore, when CNNs learned from different cells are transferred between cells, there is a diversity of transfer learning performance, which indicates that CNNs are cell-specific. Moreover, when CNNs are transferred between different types of input images, here white noise v.s. natural images, transfer learning shows a good performance, which implies that CNNs indeed capture the full computational ability of a single retinal cell for different inputs. Taken together, these results suggest that CNNs could be used to reveal structure components of neuronal circuits, and provide a powerful model for neural system identification.

q-bio.NC↗

arXiv2020

Reconstruction of Natural Visual Scenes from Neural Spikes with Deep Neural Networks

Neural coding is one of the central questions in systems neuroscience for understanding how the brain processes stimulus from the environment, moreover, it is also a cornerstone for designing algorithms of brain-machine interface, where decoding incoming stimulus is highly demanded for better performance of physical devices. Traditionally researchers have focused on functional magnetic resonance imaging (fMRI) data as the neural signals of interest for decoding visual scenes. However, our visual perception operates in a fast time scale of millisecond in terms of an event termed neural spike. There are few studies of decoding by using spikes. Here we fulfill this aim by developing a novel decoding framework based on deep neural networks, named spike-image decoder (SID), for reconstructing natural visual scenes, including static images and dynamic videos, from experimentally recorded spikes of a population of retinal ganglion cells. The SID is an end-to-end decoder with one end as neural spikes and the other end as images, which can be trained directly such that visual scenes are reconstructed from spikes in a highly accurate fashion. Our SID also outperforms on the reconstruction of visual stimulus compared to existing fMRI decoding models. In addition, with the aid of a spike encoder, we show that SID can be generalized to arbitrary visual scenes by using the image datasets of MNIST, CIFAR10, and CIFAR100. Furthermore, with a pre-trained SID, one can decode any dynamic videos to achieve real-time encoding and decoding of visual scenes by spikes. Altogether, our results shed new light on neuromorphic computing for artificial visual systems, such as event-based visual cameras and visual neuroprostheses.

q-bio.NC↗

arXiv2020

P-ODN: Prototype based Open Deep Network for Open Set Recognition

Most of the existing recognition algorithms are proposed for closed set scenarios, where all categories are known beforehand. However, in practice, recognition is essentially an open set problem. There are categories we know called "knowns", and there are more we do not know called "unknowns". Enumerating all categories beforehand is never possible, consequently it is infeasible to prepare sufficient training samples for those unknowns. Applying closed set recognition methods will naturally lead to unseen-category errors. To address this problem, we propose the prototype based Open Deep Network (P-ODN) for open set recognition tasks. Specifically, we introduce prototype learning into open set recognition. Prototypes and prototype radiuses are trained jointly to guide a CNN network to derive more discriminative features. Then P-ODN detects the unknowns by applying a multi-class triplet thresholding method based on the distance metric between features and prototypes. Manual labeling the unknowns which are detected in the previous process as new categories. Predictors for new categories are added to the classification layer to "open" the deep neural networks to incorporate new categories dynamically. The weights of new predictors are initialized exquisitely by applying a distances based algorithm to transfer the learned knowledge. Consequently, this initialization method speed up the fine-tuning process and reduce the samples needed to train new predictors. Extensive experiments show that P-ODN can effectively detect unknowns and needs only few samples with human intervention to recognize a new category. In the real world scenarios, our method achieves state-of-the-art performance on the UCF11, UCF50, UCF101 and HMDB51 datasets.

cs.CV↗

arXiv2020

Towards the Next Generation of Retinal Neuroprosthesis: Visual Computation with Spikes

Neuroprosthesis, as one type of precision medicine device, is aiming for manipulating neuronal signals of the brain in a closed-loop fashion, together with receiving stimulus from the environment and controlling some part of our brain/body. In terms of vision, incoming information can be processed by the brain in millisecond interval. The retina computes visual scenes and then sends its output as neuronal spikes to the cortex for further computation. Therefore, the neuronal signal of interest for retinal neuroprosthesis is spike. Closed-loop computation in neuroprosthesis includes two stages: encoding stimulus to neuronal signal, and decoding it into stimulus. Here we review some of the recent progress about visual computation models that use spikes for analyzing natural scenes, including static images and dynamic movies. We hypothesize that for a better understanding of computational principles in the retina, one needs a hypercircuit view of the retina, in which different functional network motifs revealed in the cortex neuronal network should be taken into consideration for the retina. Different building blocks of the retina, including a diversity of cell types and synaptic connections, either chemical synapses or electrical synapses (gap junctions), make the retina an ideal neuronal network to adapt the computational techniques developed in artificial intelligence for modeling of encoding/decoding visual scenes. Altogether, one needs a systems approach of visual computation with spikes to advance the next generation of retinal neuroprosthesis as an artificial visual system.

q-bio.NC↗

arXiv2019

An Efficient Coding Method for Spike Camera using Inter-Spike Intervals

Recently, a novel bio-inspired spike camera has been proposed, which continuously accumulates luminance intensity and fires spikes while the dispatch threshold is reached. Compared to the conventional frame-based cameras and the emerging dynamic vision sensors, the spike camera has shown great advantages in capturing fast-moving scene in a frame-free manner with full texture reconstruction capabilities. However, it is difficult to transmit or store the large amount of spike data. To address this problem, we first investigate the spatiotemporal distribution of inter-spike intervals and propose an intensity-based measurement of spike train distance. Then, we design an efficient spike coding method, which integrates the techniques of adaptive temporal partitioning, intra-/inter-pixel prediction, quantization and entropy coding into a unified lossy coding framework. Finally, we construct a PKU-Spike dataset captured by the spike camera to evaluate the compression performance. The experimental results on the dataset demonstrate that the proposed approach is effective in compressing such spike data while maintaining the fidelity.

cs.MM↗

arXiv2019

Transductive Episodic-Wise Adaptive Metric for Few-Shot Learning

Few-shot learning, which aims at extracting new concepts rapidly from extremely few examples of novel classes, has been featured into the meta-learning paradigm recently. Yet, the key challenge of how to learn a generalizable classifier with the capability of adapting to specific tasks with severely limited data still remains in this domain. To this end, we propose a Transductive Episodic-wise Adaptive Metric (TEAM) framework for few-shot learning, by integrating the meta-learning paradigm with both deep metric learning and transductive inference. With exploring the pairwise constraints and regularization prior within each task, we explicitly formulate the adaptation procedure into a standard semi-definite programming problem. By solving the problem with its closed-form solution on the fly with the setup of transduction, our approach efficiently tailors an episodic-wise metric for each task to adapt all features from a shared task-agnostic embedding space into a more discriminative task-specific metric space. Moreover, we further leverage an attention-based bi-directional similarity strategy for extracting the more robust relationship between queries and prototypes. Extensive experiments on three benchmark datasets show that our framework is superior to other existing approaches and achieves the state-of-the-art performance in the few-shot literature.

cs.LG↗

arXiv2019

Exploring Reciprocal Attention for Salient Object Detection by Cooperative Learning

Typically, objects with the same semantics are not always prominent in images containing different backgrounds. Motivated by this observation that accurately salient object detection is related to both foreground and background, we proposed a novel cooperative attention mechanism that jointly considers reciprocal relationships between background and foreground for efficient salient object detection. Concretely, we first aggregate the features at each side-out of traditional dilated FCN to extract the initial foreground and background local responses respectively. Then taking these responses as input, reciprocal attention module adaptively models the nonlocal dependencies between any two pixels of the foreground and background features, which is then aggregated with local features in a mutual reinforced way so as to enhance each branch to generate more discriminative foreground and background saliency map. Besides, cooperative losses are particularly designed to guide the multi-task learning of foreground and background branches, which encourages our network to obtain more complementary predictions with clear boundaries. At last, a simple but effective fusion strategy is utilized to produce the final saliency map. Comprehensive experimental results on five benchmark datasets demonstrate that our proposed method performs favorably against the state-of-the-art approaches in terms of all compared evaluation metrics.

cs.CV↗

arXiv2019

Selectivity or Invariance: Boundary-aware Salient Object Detection

Typically, a salient object detection (SOD) model faces opposite requirements in processing object interiors and boundaries. The features of interiors should be invariant to strong appearance change so as to pop-out the salient object as a whole, while the features of boundaries should be selective to slight appearance change to distinguish salient objects and background. To address this selectivity-invariance dilemma, we propose a novel boundary-aware network with successive dilation for image-based SOD. In this network, the feature selectivity at boundaries is enhanced by incorporating a boundary localization stream, while the feature invariance at interiors is guaranteed with a complex interior perception stream. Moreover, a transition compensation stream is adopted to amend the probable failures in transitional regions between interiors and boundaries. In particular, an integrated successive dilation module is proposed to enhance the feature invariance at interiors and transitional regions. Extensive experiments on six datasets show that the proposed approach outperforms 16 state-of-the-art methods.

cs.CV↗

arXiv2019

Distortion-adaptive Salient Object Detection in 360$^\circ$ Omnidirectional Images

Image-based salient object detection (SOD) has been extensively explored in the past decades. However, SOD on 360$^\circ$ omnidirectional images is less studied owing to the lack of datasets with pixel-level annotations. Toward this end, this paper proposes a 360$^\circ$ image-based SOD dataset that contains 500 high-resolution equirectangular images. We collect the representative equirectangular images from five mainstream 360$^\circ$ video datasets and manually annotate all objects and regions over these images with precise masks with a free-viewpoint way. To the best of our knowledge, it is the first public available dataset for salient object detection on 360$^\circ$ scenes. By observing this dataset, we find that distortion from projection, large-scale complex scene and small salient objects are the most prominent characteristics. Inspired by these foundings, this paper proposes a baseline model for SOD on equirectangular images. In the proposed approach, we construct a distortion-adaptive module to deal with the distortion caused by the equirectangular projection. In addition, a multi-scale contextual integration block is introduced to perceive and distinguish the rich scenes and objects in omnidirectional scenes. The whole network is organized in a progressively manner with deep supervision. Experimental results show the proposed baseline approach outperforms the top-performanced state-of-the-art methods on 360$^\circ$ SOD dataset. Moreover, benchmarking results of the proposed baseline approach and other methods on 360$^\circ$ SOD dataset show the proposed dataset is very challenging, which also validate the usefulness of the proposed dataset and approach to boost the development of SOD on 360$^\circ$ omnidirectional scenes.

cs.CV↗

arXiv2019

Learning Local Feature Descriptor with Motion Attribute for Vision-based Localization

In recent years, camera-based localization has been widely used for robotic applications, and most proposed algorithms rely on local features extracted from recorded images. For better performance, the features used for open-loop localization are required to be short-term globally static, and the ones used for re-localization or loop closure detection need to be long-term static. Therefore, the motion attribute of a local feature point could be exploited to improve localization performance, e.g., the feature points extracted from moving persons or vehicles can be excluded from these systems due to their unsteadiness. In this paper, we design a fully convolutional network (FCN), named MD-Net, to perform motion attribute estimation and feature description simultaneously. MD-Net has a shared backbone network to extract features from the input image and two network branches to complete each sub-task. With MD-Net, we can obtain the motion attribute while avoiding increasing much more computation. Experimental results demonstrate that the proposed method can learn distinct local feature descriptor along with motion attribute only using an FCN, by outperforming competing methods by a wide margin. We also show that the proposed algorithm can be integrated into a vision-based localization algorithm to improve estimation accuracy significantly.

cs.CV↗

arXiv2019

A Retina-inspired Sampling Method for Visual Texture Reconstruction

Conventional frame-based camera is not able to meet the demand of rapid reaction for real-time applications, while the emerging dynamic vision sensor (DVS) can realize high speed capturing for moving objects. However, to achieve visual texture reconstruction, DVS need extra information apart from the output spikes. This paper introduces a fovea-like sampling method inspired by the neuron signal processing in retina, which aims at visual texture reconstruction only taking advantage of the properties of spikes. In the proposed method, the pixels independently respond to the luminance changes with temporal asynchronous spikes. Analyzing the arrivals of spikes makes it possible to restore the luminance information, enabling reconstructing the natural scene for visualization. Three decoding methods of spike stream for texture reconstruction are proposed for high-speed motion and stationary scenes. Compared to conventional frame-based camera and DVS, our model can achieve better image quality and higher flexibility, which is capable of changing the way that demanding machine vision applications are built.

eess.IV↗

arXiv2019

2D LiDAR Map Prediction via Estimating Motion Flow with GRU

It is a significant problem to predict the 2D LiDAR map at next moment for robotics navigation and path-planning. To tackle this problem, we resort to the motion flow between adjacent maps, as motion flow is a powerful tool to process and analyze the dynamic data, which is named optical flow in video processing. However, unlike video, which contains abundant visual features in each frame, a 2D LiDAR map lacks distinctive local features. To alleviate this challenge, we propose to estimate the motion flow based on deep neural networks inspired by its powerful representation learning ability in estimating the optical flow of the video. To this end, we design a recurrent neural network based on gated recurrent unit, which is named LiDAR-FlowNet. As a recurrent neural network can encode the temporal dynamic information, our LiDAR-FlowNet can estimate motion flow between the current map and the unknown next map only from the current frame and previous frames. A self-supervised strategy is further designed to train the LiDAR-FlowNet model effectively, while no training data need to be manually annotated. With the estimated motion flow, it is straightforward to predict the 2D LiDAR map at the next moment. Experimental results verify the effectiveness of our LiDAR-FlowNet as well as the proposed training strategy. The results of the predicted LiDAR map also show the advantages of our motion flow based method.

cs.RO↗

arXiv2019

ODN: Opening the Deep Network for Open-set Action Recognition

In recent years, the performance of action recognition has been significantly improved with the help of deep neural networks. Most of the existing action recognition works hold the \textit{closed-set} assumption that all action categories are known beforehand while deep networks can be well trained for these categories. However, action recognition in the real world is essentially an \textit{open-set} problem, namely, it is impossible to know all action categories beforehand and consequently infeasible to prepare sufficient training samples for those emerging categories. In this case, applying closed-set recognition methods will definitely lead to unseen-category errors. To address this challenge, we propose the Open Deep Network (ODN) for the open-set action recognition task. Technologically, ODN detects new categories by applying a multi-class triplet thresholding method, and then dynamically reconstructs the classification layer and "opens" the deep network by adding predictors for new categories continually. In order to transfer the learned knowledge to the new category, two novel methods, Emphasis Initialization and Allometry Training, are adopted to initialize and incrementally train the new predictor so that only few samples are needed to fine-tune the model. Extensive experiments show that ODN can effectively detect and recognize new categories with little human intervention, thus applicable to the open-set action recognition tasks in the real world. Moreover, ODN can even achieve comparable performance to some closed-set methods.

cs.CV↗

arXiv2018

Hierarchical Deep Co-segmentation of Primary Objects in Aerial Videos

Primary object segmentation plays an important role in understanding videos generated by unmanned aerial vehicles. In this paper, we propose a large-scale dataset with 500 aerial videos and manually annotated primary objects. To the best of our knowledge, it is the largest dataset to date for primary object segmentation in aerial videos. From this dataset, we find most aerial videos contain large-scale scenes, small primary objects as well as consistently varying scales and viewpoints. Inspired by that, we propose a hierarchical deep co-segmentation approach that repeatedly divides a video into two sub-videos formed by the odd and even frames, respectively. In this manner, the primary objects shared by sub-videos can be co-segmented by training two-stream CNNs and finally refined within the neighborhood reversible flows. Experimental results show that our approach remarkably outperforms 17 state-of-the-art methods in segmenting primary objects in various types of aerial videos.

cs.CV↗