Hand detection is a crucial pre-processing procedure for many human hand related computer vision tasks, such as hand pose estimation, hand gesture recognition, human activity analysis, and so on. However, reliably detecting multiple hands from cluttering scenes remains to be a challenging task because of complex appearance diversities of dexterous human hands (e.g., different hand shapes, skin colors, illuminations, orientations, and scales, etc.) in color images. To tackle this problem, an accurate hand detection method is proposed to reliably detect multiple hands from a single color image using a hybrid detection/reconstruction convolutional neural networks (CNN) framework, in which regions of hands are detected and appearances of hands are reconstructed in parallel by sharing features extracted from a region proposal layer, and the proposed model is trained in an end-to-end manner. Furthermore, it is observed that the generative adversarial network (GAN) could further boost the detection performance by generating more realistic hand appearances. The experimental results show that the proposed approach outperforms the state-of-the-art on public challenging hand detection benchmarks.

Anchor-free detection methods identify different objects by perceiving bounding box keypoints without predefined anchor boxes, which have attracted much attention due to their straightforward design and comparable performance. Currently, most anchor-free methods detect bounding box corners to regress object locations. In clutter environments, the bounding box corners may lie in background regions, which have limited relation with the object itself. In addition, the relationships between object keypoints are always neglected, potentially affecting the perceptibility of the detector for high-precision object detection. In this paper, we propose the Keypoint Relational Regression Network (KRRNet) to detect object keypoints with semantic relations instead of bounding box corners. The relational regression head is designed to enhance the keypoint relationship exploration capability and reason accuracy object locations. Moreover, the random background sampling strategy is proposed to sample negative background points around foreground object regions and form point pairs with object keypoints. Then, KRRNet can explicitly learn discriminative feature embedding from contrastive learning to pull close the positive pairs and push apart the negative pairs, resisting the influence of surrounding complex environments. KRRNet can be trained on one Nvidia RTX 3090 GPU and achieves a single-scale test AP of 48.9% and multi-scale test AP of 50.6% on the MS-COCO test-dev with the backbone of Hourglass-104, surpassing state-of-the-art bottom-up anchor-free detector using the same backbone.