Beyond MOT: Semantic Multi-Object Tracking

Current multi-object tracking (MOT) aims to predict trajectories of targets (i.e.,"where") in videos. Yet, knowing merely "where" is insufficient in many crucial applications. In comparison, semantic understanding such as fine-grained behaviors, interactions, and overall summarized captions (i.e., "what") from videos, associated with "where", is highly-desired for comprehensive video analysis. Thus motivated, we introduce Semantic Multi-Object Tracking (SMOT), that aims to estimate object trajectories and meanwhile understand semantic details of associated trajectories including instance captions, instance interactions, and overall video captions, integrating "where" and "what" for tracking. In order to foster the exploration of SMOT, we propose BenSMOT, a large-scale Benchmark for Semantic MOT. Specifically, BenSMOT comprises 3,292 videos with 151K frames, covering various scenarios for semantic tracking of humans. BenSMOT provides annotations for the trajectories of targets, along with associated instance captions in natural language, instance interactions, and overall caption for each video sequence. To our best knowledge, BenSMOT is the first publicly available benchmark for SMOT. Besides, to encourage future research, we present a novel tracker named SMOTer, which is specially designed and end-to-end trained for SMOT, showing promising performance. By releasing BenSMOT, we expect to go beyond conventional MOT by predicting "where" and "what" for SMOT, opening up a new direction in tracking for video understanding. Our BenSMOT and SMOTer will be released.

TSViT: A Time Series Vision Transformer for Fault Diagnosis

Traditional fault diagnosis methods using Convolutional Neural Networks (CNNs) face limitations in capturing temporal features (i.e., the variation of vibration signals over time). To address this issue, this paper introduces a novel model, the Time Series Vision Transformer (TSViT), specifically designed for fault diagnosis. On one hand, TSViT model integrates a convolutional layer to segment vibration signals and capture local features. On the other hand, it employs a transformer encoder to learn long-term temporal information. The experimental results with other methods on two distinct datasets validate the effectiveness and generalizability of TSViT with a comparative analysis of its hyperparameters' impact on model performance, computational complexity, and overall parameter quantity. TSViT reaches average accuracies of 100% and 99.99% on two test sets, correspondingly.