TL;DR
This work introduces a Tri-Directional Attention module into the backbone of the YOLOv11 baseline framework, modeling the spatial position information via a three-way attention mechanism to capture richer fine-grained features and devise a Cross-Dimension Feature Interaction module to enhance the recognition accuracy for non-salient objects.
Enhanced Feature-Aware YOLO (EFA-YOLO), a lightweight detection framework based on YOLO11n, achieves a favorable balance between detection accuracy and computational efficiency, demonstrating real-time performance on the evaluated GPU platform and indicating potential for further edge-oriented optimization.
UAV imagery is particularly difficult for object detection because targets occupy few pixels, occur in dense or occluded groups, and must be separated from extensive background texture. Adding a high-resolution detection layer improves the visibility of tiny instances, but it also carries road markings, building contou...
MFP-YOLO, a lightweight detector based on YOLOv11, is developed by introducing targeted improvements in feature extraction, feature fusion, and bounding-box regression, demonstrating that the integrated modifications provide a favorable trade-off between detection accuracy and computational efficiency for remote-sensin...
A novel feature learning framework named DiffReID is proposed for object ReID that leverages a discriminative diffusion model to gradually learn identity-aware distributions and generate identity-invariant features.
Tests on the COCO dataset show that EnYOLO-World offers a better trade-off between accuracy and speed compared to existing real-time detectors, demonstrating the value of optimizing both backbone feature retention and neck-level fusion strategies.
The proposed Multi-scale Attention Enhancement Feature Fusion Network (MAEF-Net) systematically optimizes the overall detection pipeline from three aspects, namely feature enhancement, feature fusion, and prediction refinement, and achieves a favorable balance between detection accuracy and computational efficiency whi...
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
