Graph convolutional networks (GCNs) have attracted considerable interest in skeleton-based action recognition. Existing GCN-based models have proposed methods to learn dynamic graph topologies generated from the feature information of vertices to capture inherent relationships. However, these models have two main limitations. Firstly, they struggle to effectively utilize high-dimensional or structural information, which limits their capacity for feature representation and consequently hinders performance improvement. Secondly, among these models, the multi-scale methods that aggregate information at different scales often over-capture unnecessary relationships between vertices. This leads to an over-smoothing problem where smoothed features are extracted, making it difficult to distinguish the features of each vertex. To address these limitations, we propose the multi-scale structural graph convolutional network (MSS-GCN) for skeleton-based action recognition. Within the MSS-GCN framework, the common intersection graph convolution (CI-GC) leverages the overlapped neighbor information, indicating the overlap between neighboring vertices for a given pair of root vertices. The graph topology of CI-GC is designed to compute the structural correlation between neighboring vertices corresponding to each hop, thereby enriching the context of inter-vertex relationships. Then, our proposed multi-scale spatio-temporal modeling aggregates local-global features to provide a comprehensive representation. In addition, we propose a Graph Weight Annealing (GWA) method, which is a graph scheduling method to mitigate the over-smoothing caused by multi-scale aggregation. By varying the importance between a vertex and its neighbors, we demonstrate that the over-smoothing problem can be effectively mitigated. Moreover, our proposed GWA method can easily be adapted to different GCN models to enhance performance. Combining the MSS-GCN model and the GWA method, we propose a powerful feature extractor that effectively classifies actions for skeleton-based action recognition in various datasets. We evaluate our approach on three benchmark datasets: NTU RGB+D, NTU RGB+D 120, and NW-UCLA. The proposed MSS-GCN achieves state-of-the-art performance on all three datasets, further validating the effectiveness of our approach.

Multi-object tracking (MOT), which is crucial for computer vision and video processing, has immense potential for improvement. Traditional tracking-by-detection approaches include feature-based object re-identification methods that use trained features, but these methods suffer from a lack of suitable training data. In training datasets used for MOT, every object in a video sequence must have its own location and ID. However, assigning IDs to each object in every sequence is considerably labor-intensive, and hence current MOT datasets are unsuitable for training re-identification networks. To resolve this issue, this paper proposes a novel self-supervised learning method using several short videos that contain no human-added labels, based on the idea that each video is a set of temporally corresponding image frames. We then describe how to improve tracking performance using a re-identification network trained in a self-supervised manner. In addition, ablation studies were conducted in order to define the optimal parameters, such as number of clips, data augmentation, and appropriate matching algorithms. The proposed approach achieved competitive performance compared with current best-practice methods including supervised methods, achieving MOT accuracy = 62.0% and ID F1-score = 62.7% on the MOT17 benchmark.

Multi-person pose estimation has been gaining considerable interest due to its use in several real-world applications, such as activity recognition, motion capture, and augmented reality. Although the improvement of the accuracy and speed of multi-person pose estimation techniques has been recently studied, limitations still exist in balancing these two aspects. In this paper, a novel knowledge distilled lightweight top-down pose network (KDLPN) is proposed that balances computational complexity and accuracy. For the first time in multi-person pose estimation, a network that reduces computational complexity by applying a “Pelee” structure and shuffles pixels in the dense upsampling convolution layer to reduce the number of channels is presented. Furthermore, to prevent performance degradation because of the reduced computational complexity, knowledge distillation is applied to establish the pose estimation network as a teacher network. The method performance is evaluated on the MSCOCO dataset. Experimental results demonstrate that our KDLPN network significantly reduces 95% of the parameters required by state-of-the-art methods with minimal performance degradation. Moreover, our method is compared with other pose estimation methods to substantiate the importance of computational complexity reduction and its effectiveness.

Video-based person re-identification (ReID) aims to exploit relevant features from spatial and temporal knowledge. Widely used methods include the part- and attention-based approaches for suppressing irrelevant spatial–temporal features. However, it is still challenging to overcome inconsistencies across video frames due to occlusion and imperfect detection. These mismatches make temporal processing ineffective and create an imbalance of crucial spatial information. To address these problems, we propose the Spatiotemporal Multi-Granularity Aggregation (ST-MGA) method, which is specifically designed to accumulate relevant features with spatiotemporally consistent cues. The proposed framework consists of three main stages: extraction, which extracts spatiotemporally consistent partial information; augmentation, which augments the partial information with different granularity levels; and aggregation, which effectively aggregates the augmented spatiotemporal information. We first introduce the consistent part-attention (CPA) module, which extracts spatiotemporally consistent and well-aligned attentive parts. Sub-parts derived from CPA provide temporally consistent semantic information, solving misalignment problems in videos due to occlusion or inaccurate detection, and maximize the efficiency of aggregation through uniform partial information. To enhance the diversity of spatial and temporal cues, we introduce the Multi-Attention Part Augmentation (MA-PA) block, which incorporates fine parts at various granular levels, and the Long-/Short-term Temporal Augmentation (LS-TA) block, designed to capture both long- and short-term temporal relations. Using densely separated part cues, ST-MGA fully exploits and aggregates the spatiotemporal multi-granular patterns by comparing relations between parts and scales. In the experiments, the proposed ST-MGA renders state-of-the-art performance on several video-based ReID benchmarks (i.e., MARS, DukeMTMC-VideoReID, and LS-VID).