Yipeng Li

Recent years have witnessed remarkable achievements in video-based action recognition. Apart from traditional frame-based cameras, event cameras are bio-inspired vision sensors that only record pixel-wise brightness changes rather than the brightness value. However, little effort has been made in event-based action recognition, and large-scale public datasets are also nearly unavailable. In this paper, we propose an event-based action recognition framework called EV-ACT . The Learnable Multi-Fused Representation (LMFR) is first proposed to integrate multiple event information in a learnable manner. The LMFR with dual temporal granularity is fed into the event-based slow-fast network for the fusion of appearance and motion features. A spatial-temporal attention mechanism is introduced to further enhance the learning capability of action recognition. To prompt research in this direction, we have collected the largest event-based action recognition benchmark named THU^E-ACT-50 and the accompanying THU^E-ACT-50-CHL dataset under challenging environments, including a total of over 12,830 recordings from 50 action categories, which is over 4 times the size of the previous largest dataset. Experimental results show that our proposed framework could achieve improvements of over 14.5%, 7.6%, 11.2%, and 7.4% compared to previous works on four benchmarks. We have also deployed our proposed EV-ACT framework on a mobile platform to validate its practicality and efficiency.

Multiple image hiding aims to hide multiple secret images into a single cover image, and then recover all secret images perfectly. Such high-capacity hiding may easily lead to contour shadows or color distortion, which makes multiple image hiding a very challenging task. In this paper, we propose a novel multiple image hiding framework based on invertible neural network, namely DeepMIH. Specifically, we develop an invertible hiding neural network (IHNN) to innovatively model the image concealing and revealing as its forward and backward processes, making them fully coupled and reversible. The IHNN is highly flexible, which can be cascaded as many times as required to achieve the hiding of multiple images. To enhance the invisibility, we design an importance map (IM) module to guide the current image hiding based on the previous image hiding results. In addition, we find that the image hidden in the high-frequency sub-bands tends to achieve better hiding performance, and thus propose a low-frequency wavelet loss to constrain that no secret information is hidden in the low-frequency sub-bands. Experimental results show that our DeepMIH significantly outperforms other state-of-the-art methods, in terms of hiding invisibility, security and recovery accuracy on a variety of datasets.

According to the application requirements of microwave broadband information distribution and mutual positioning in warship formation combat system, an integrated communication and localization method based on digital phased array antenna is proposed, in which the simultaneous multi-beam forming technology is applied to shorten the time of signal search and communication link building, the monopulse sum-difference beam approach and two-way transmission delay are employed correspondingly to carry out accurate direction finding and ranging, moreover, in view of correlation characteristics of communication signal in time domain, the filtering operation is conducted so as to further improve the positioning performance. The communication and localization function are designed unitarily in the method presented in this paper, which can reduce the occupied space on warship platform and enhance stealth and electromagnetic compatibility performance. What’s more, it is not dependent on satellite system, especially suitable for cooperative combat occasions of warship formation under weak navigation or even satellite rejection environment.

Most multi-view based human pose estimation techniques assume the cameras are fixed. While in dynamic scenes, the cameras should be able to move and seek the best views to avoid occlusions and extract 3D information of the target collaboratively. In this paper, we address the problem of online view selection for a fixed number of cameras to estimate multi-person 3D poses actively. The proposed method exploits a distributed multi-agent based deep reinforcement learning framework, where each camera is modeled as an agent, to optimize the action of all the cameras. An inter-agent communication protocol was developed to transfer the cameras’ relative positions between agents for better collaboration. Experiments on the Panoptic dataset show that our method outperforms other view selection methods by a large margin given an identical number of cameras. To the best of our knowledge, our method is the first to address online active multi-view 3D pose estimation with multi-agent reinforcement learning.

It is widely acknowledged that biological intelligence is capable of learning continually without forgetting previously learned skills. Unfortunately, it has been widely observed that many artificial intelligence techniques, especially (deep) neural network (NN)-based ones, suffer from catastrophic forgetting problem, which severely forgets previous tasks when learning a new one. How to train NNs without catastrophic forgetting, which is termed continual learning, is emerging as a frontier topic and attracting considerable research interest. Inspired by memory replay and synaptic consolidation mechanism in brain, in this article, we propose a novel and simple framework termed memory recall (MeRec) for continual learning with deep NNs. In particular, we first analyze the feature stability across tasks in NN and show that NN can yield task stable features in certain layers. Then, based on this observation, we use a memory module to keep the feature statistics (mean and std) for each learned task. Based on the memory and statistics, we show that a simple replay strategy with Gaussian distribution-based feature regeneration can recall and recover the knowledge from previous tasks. Together with the weight regularization, MeRec preserves weights learned from previous tasks. Based on this simple framework, MeRec achieved leading performance with extremely small memory budget (only two feature vectors for each class) for continual learning on CIFAR-10 and CIFAR-100 datasets, with at least 50% accuracy drop reduction after several tasks compared to previous state-of-the-art approaches.