RePre: Improving Self-Supervised Vision Transformer with Reconstructive Pre-training

Abstract

Corner-guided detector enjoys potential ability to yield precise bounding boxes. However, unreliable corner pairs, generated by heuristic grouping guidance, hinder the development of this detector. In this paper, we propose a novel corner grouping algorithm, termed as Corner Affinity, to significantly boost the reliability and robustness of corner grouping. The proposed Corner Affinity is a couple of two interactional factors, namely, 1) the structure affinity (SA), applying to generate preliminary corner pairs through the corresponding object's shallow construction information. 2) the contexts affinity (CA), running as optimizing corner pairs via embedding deeper semantic features of affiliated instances. Equipped with the Corner Affinity, a detector can produce high-quality bounding boxes upon preferable paired corner keypoints. Experimental results show the superiority of our design on multiple benchmark datasets. Specifically, for CornerNet baseline, the proposed Corner Affinity brings AP boostings of 5.8% on COCO, 35.8% on Citypersons, and 17.2% on UCAS-AOD without bells and whistles.

Wearable-based behaviour interpolation for semi-supervised human activity recognition

“…It is well-known that body-worn sensors are used for many real-world applications, including but not limited to sleep monitoring [1], environmental perception [2], elderly patient assistance and health assessment [3], etc. Also, wearable-based human activity recognition (HAR) is one of the core research areas in ubiquitous computing, and it plays an essential role in human behaviour understanding, health monitoring, skill assessment, sports training, etc.…”

“…Fig. 1 shows the HAR model performance of different learning paradigms, and it seems that the feature learned by recent self-supervised work 1 [5] on the wearable-based dataset does not help much when compared with supervised training.…”