Motion Recognition Algorithm Based on Deep Edge-Aware Pyramid Pooling Network in Human–Computer Interaction

Abstract

Action recognition is an important research direction in computer vision, which has worldwide applications, such as video surveillance, human-robot interaction and so on. Due to the influence of complex background and multi-angle changes, accurate recognition and analysis of human motion in real-life scenarios is still a challenging problem. In order to improve the accuracy of pedestrian detection and motion recognition, this paper proposes a novel edge-aware end-to-end deep network method, which uses the edge-aware pooling module to improve pedestrian contour accuracy and captures video sequences using multi-scale pyramid pooling layer spatial-time context feature. The complementary features of the edge-related features can effectively preserve the clear boundary, and the combination of the auxiliary side output and the pyramid pooling layer output can extract rich global context information. A large number of qualitative and quantitative experimental results show that the proposed model can effectively improve the performance of existing pedestrian detection and motion recognition networks on the UCF-101, HMDB-51, and KTH dataset. INDEX TERMS Motion recognition, edge perception, deep learning, pyramid pooling, spatial-temporal context.

Contrastive Semantic‐Guided Image Smoothing Network

“…Data‐driven methods : Recently, many deep learning‐based methods have been proposed and achieved remarkable performance. Xu et al [XRY*15] propose a deep edge‐aware filter to approximate different operators based on a gradient domain training procedure. Shen et al [SCTJ17] present a pyramid structure with an extended receptive field.…”

“…With the development of deep learning, data‐driven approaches [XRY*15, ZLJ*19] have also been explored on image smoothing. The deep hierarchical network structure has a natural advantage in capturing semantic cues for extracting meaningful image structures.…”

Robust detection method for improving small traffic sign recognition based on spatial pyramid pooling

“…SPP is a manageable answer for managing aspect ratios, diverse scales, dimensions, and sizes (Wang et al 2018;Fu et al 2020). These results are significant in computer visual recognition but have gained little attention in deep network research (Xu et al 2019a). Figure 1 presents a network configuration with an SPP layer.…”

Recognising pedestrian behaviour using a multi-channel spatiotemporal fusion network

“…However, the long and short-term memory network in the recurrent neural network is typically used to solve the correlation problem between consecutive frames in the video, as only taking into account the CNN will ignore the continuity problem of image frames. Accordingly, [9] proposes a sort of end-to-end fullyconnected long and short-term memory (LSTM) network for the human behaviour recognition technology based on human behaviour; For example, [10] applies video recognition technology for the first time by combining a CNN with a long-and short-term memory network; [11][12][13] combines the two networks and uses the dual-stream network to extract features in the time and space dimensions, respectively, to achieve good recognition results. blended and used in the field of video recognition technology; [11][12][13] successfully recognised human behaviour by combining the two networks and using a dual-stream network to extract features in the temporal and spatial dimensions, respectively.…”