Multi-person eyeblink detection in untrimmed in-the-wild videos is a recently emerged and challenging task. Due to its significant spatio-temporal fine-grained characteristics compared to general actions, we empirically find that general action detectors, though effective in general domains, struggle with this task (i.e., Blink-AP < 2%). Specialized eyeblink detection methods alleviate it through fine-grained spatio-temporal operations. SOTA method proposes a unified model combining instance-aware face localization and eyeblink detection through joint multi-task learning and feature sharing. While effective, it exhibits two critical limitations that may contribute to its unsatisfactory performance (i.e., Blink-AP=10.11%): (1) Face localization and eyeblink detection require distinct spatio-temporal feature granularities, making joint modeling in a unified feature space suboptimal. (2) Eyeblink task training could be largely affected by unstable face-eye feature learning under the joint training paradigm. To address this, we propose DeFB, a decomposed feature learning paradigm with favorable effectiveness and efficiency: (1) We model faces and eyes in granularity-specific feature spaces, which enhances fine-grained perception while reducing computational costs compared to a unified feature space. (2) To mitigate face-eye feature learning instability, we adopt an asynchronous learning mechanism where eye feature learning refines well-trained coarse face features, with shared queries acting as a bridge between stages to retain the efficient feature sharing of existing unified models. Compared with SOTA method, DeFB doubles the performance (Blink-AP: 24.65% v.s. 10.11%) while boosting efficiency by nearly 35%. DeFB can also be integrated as a plug-in to substantially augment the eyeblink detection capabilities of general action detectors.

Designing a novel and efficient architecture is the thrust of medical image segmentation. In this paper, we introduce a novel network named Multi-scale Feedback Feature Refinement U-Net (MFFRU-Net) for medical image segmentation. We design a simple and effective multi-scale feedback structure. Up-sampling and 1 × 1 convolution are used to feedback the feature maps of different scales in the decoder to the encoder, so that multiple high-level and low-level features are fused to obtain more representative features. Specifically, we propose a feature refinement module (FRM) based on the dual attention mechanism in the middle layer of the network. FRM block can enhance the use of spatial and channel information of image features. We evaluate the MFFRU-Net on two datasets. Comprehensive experimental results show that the proposed method is superior to the original U-Net method and other advanced methods.