Yongliang Zhang
Yongliang Zhang has a total of 4 co-authors. Here are the top co-authors by number of shared publications:
- Nan Geng
College of Information Engineering, Northwest A&F University, Xianyang 712100, China
2 shared publications - Yue Yang
College of Information Engineering, Northwest A&F University, Xianyang 712100, China
2 shared publications - Keyuan Wang
College of Information Engineering, Northwest A&F University, Xianyang 712100, China
1 shared publication - Ronggeng Guo
College of Information Engineering, Northwest A&F University, Xianyang 712100, China
1 shared publication
Latest Publications
- Open Access
DRG-LiteStar-YOLO: Reliability-Guided Pseudo-Depth Fusion for Dairy Goat Detection in Complex Barns
Reliable dairy goat detection in barns is challenging because uneven illumination, railings, and animal overlap weaken RGB boundaries. We developed DRG-LiteStar-YOLO, a lightweight RGB-camera-compatible detector that integrates monocular pseudo-depth with RGB features without requiring a dedicated depth sensor. DA3-Small was used to generate relative pseudo-depth maps, which were fused with RGB features through reliability-guided fusion and boundary-enhanced multi-scale aggregation. The dataset comprised 1521 images of 134 lactating Saanen goats and 13,985 annotated instances. On the held-out test set, DRG-LiteStar-YOLO achieved a precision of 0.961, a recall of 0.941, an mAP@0.5 of 0.978, and an mAP@0.5:0.95 of 0.735. Compared with RGB-only LiteStar-YOLO, the proposed method improved mAP@0.5 and mAP@0.5:0.95 by 1.6 and 3.8 percentage points, respectively. The five-seed cumulative ablation further showed that the full configuration achieved mAP@0.5:0.95 compared with for RGB-only LiteStar-YOLO. The detector contains 4.20 M parameters and 12.60 GFLOPs and achieves 86.2 FPS on an RTX 4090 with precomputed pseudo-depth maps. These results demonstrate that reliability-guided pseudo-depth fusion improves goat localization under complex barn conditions while retaining a compact detector design. DRG-LiteStar-YOLO provides an effective perception framework for RGB-camera-based dairy goat monitoring and supports future counting, tracking, and behavior-analysis applications.Open AccessDRG-LiteStar-YOLO: Reliability-Guided Pseudo-Depth Fusion for Dairy Goat Detection in Complex Barns
Reliable dairy goat detection in barns is challenging because uneven illumination, railings, and animal overlap weaken RGB boundaries. We developed DRG-LiteStar-YOLO, a lightweight RGB-camera-compatible detector that integrates monocular pseudo-depth with RGB features without requiring a dedicated depth sensor. DA3-Small was used to generate relative pseudo-depth maps, which were fused with RGB features through reliability-guided fusion and boundary-enhanced multi-scale aggregation. The dataset comprised 1521 images of 134 lactating Saanen goats and 13,985 annotated instances. On the held-out test set, DRG-LiteStar-YOLO achieved a precision of 0.961, a recall of 0.941, an mAP@0.5 of 0.978, and an mAP@0.5:0.95 of 0.735. Compared with RGB-only LiteStar-YOLO, the proposed method improved mAP@0.5 and mAP@0.5:0.95 by 1.6 and 3.8 percentage points, respectively. The five-seed cumulative ablation further showed that the full configuration achieved mAP@0.5:0.95 compared with for RGB-only LiteStar-YOLO. The detector contains 4.20 M parameters and 12.60 GFLOPs and achieves 86.2 FPS on an RTX 4090 with precomputed pseudo-depth maps. These results demonstrate that reliability-guided pseudo-depth fusion improves goat localization under complex barn conditions while retaining a compact detector design. DRG-LiteStar-YOLO provides an effective perception framework for RGB-camera-based dairy goat monitoring and supports future counting, tracking, and behavior-analysis applications. - Open Access
- 22 May 2026
- Published by MDPI AG in Mathematics
- Vol. 14 (11) , 1786
RF-GoatDet: An Occlusion-Aware Instance Segmentation Framework for Dairy Goats with an Adaptive Receptive Field Encoder
Accurate instance segmentation is essential for precision livestock farming, as it supports individual tracking, posture analysis, body-shape measurement, and other downstream visual monitoring tasks. However, dairy goat segmentation in dense barn scenes remains challenging because frequent mutual occlusion, instance adhesion, partial visibility, and non-rigid posture variation often lead to incomplete masks and ambiguous instance boundaries. To address these challenges, this study develops RF-GoatDet, a real-time instance segmentation framework for dairy goats built upon RT-DETR. The main component of the proposed framework is an Adaptive Receptive Field Encoder (ARFE), which enhances feature encoding by adapting the effective receptive field to irregular goat contours, scale variation, and partially visible body regions. In addition, Coordinate Attention is introduced to strengthen direction-aware spatial representation, while a Query-Conditioned Dynamic Mask Head is used to generate instance-specific masks and improve the separation of adjacent goats. A dairy goat instance segmentation dataset containing 2288 annotated images was constructed, and a two-stage cleaning procedure was applied to reduce redundancy and visual anomalies. Experimental results show that RF-GoatDet achieves 46.5% mask AP and 62 FPS on this dataset, improving the RT-DETR baseline by 4.2 percentage points in mask AP while maintaining real-time inference. These results demonstrate that the proposed ARFE-centered framework effectively improves mask quality and instance discrimination in dense dairy goat scenes, providing a robust and efficient solution for real-time visual monitoring in precision livestock farming.
- 22 May 2026
- Published by MDPI AG in Mathematics
- Vol. 14 (11) , 1786
Open AccessRF-GoatDet: An Occlusion-Aware Instance Segmentation Framework for Dairy Goats with an Adaptive Receptive Field Encoder
Accurate instance segmentation is essential for precision livestock farming, as it supports individual tracking, posture analysis, body-shape measurement, and other downstream visual monitoring tasks. However, dairy goat segmentation in dense barn scenes remains challenging because frequent mutual occlusion, instance adhesion, partial visibility, and non-rigid posture variation often lead to incomplete masks and ambiguous instance boundaries. To address these challenges, this study develops RF-GoatDet, a real-time instance segmentation framework for dairy goats built upon RT-DETR. The main component of the proposed framework is an Adaptive Receptive Field Encoder (ARFE), which enhances feature encoding by adapting the effective receptive field to irregular goat contours, scale variation, and partially visible body regions. In addition, Coordinate Attention is introduced to strengthen direction-aware spatial representation, while a Query-Conditioned Dynamic Mask Head is used to generate instance-specific masks and improve the separation of adjacent goats. A dairy goat instance segmentation dataset containing 2288 annotated images was constructed, and a two-stage cleaning procedure was applied to reduce redundancy and visual anomalies. Experimental results show that RF-GoatDet achieves 46.5% mask AP and 62 FPS on this dataset, improving the RT-DETR baseline by 4.2 percentage points in mask AP while maintaining real-time inference. These results demonstrate that the proposed ARFE-centered framework effectively improves mask quality and instance discrimination in dense dairy goat scenes, providing a robust and efficient solution for real-time visual monitoring in precision livestock farming.