Xiang Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Xiang Li has a total of 68 co-authors. Here are the top co-authors by number of shared publications:
Hao Ye
College of Information and Electrical Engineering, China Agricultural University, Beijing, China
8 shared publicationsXiaonan Hu
6 shared publicationsXin Chen
6 shared publications- Xin Chen
China Agricultural University
6 shared publications - Yunsong Jia
Network Technology Center, China Agricultural University, Beijing, China
6 shared publications Xingqi Lyu
China Agricultural University
5 shared publications
Latest Publications
- 14 November 2026
- Published by Springer Nature
- p. 489-503
IMAC-AgriVLN: Can Agricultural Vision-and-Language Navigation Agents be Aware of Instruction Mistakes?
- 14 November 2026
- Published by Springer Nature
- p. 489-503
IMAC-AgriVLN: Can Agricultural Vision-and-Language Navigation Agents be Aware of Instruction Mistakes?
- 25 September 2026
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 23 (6) , 188
A lightweight YOLOv11n-based framework for accurate maize seedling detection and counting in field environments
- 25 September 2026
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 23 (6) , 188
A lightweight YOLOv11n-based framework for accurate maize seedling detection and counting in field environments
- Open Access
- 19 July 2026
- Published by Springer Nature
- p. 80-91
T-araVLN: Translator for Agricultural Robotic Agents on Vision-and-Language Navigation
- 19 July 2026
- Published by Springer Nature
- p. 80-91
Open AccessT-araVLN: Translator for Agricultural Robotic Agents on Vision-and-Language Navigation
- 19 July 2026
- Published by Springer Nature
- p. 500-511
MDE-AgriVLN: Agricultural Vision-and-Language Navigation with Monocular Depth Estimation
- 19 July 2026
- Published by Springer Nature
- p. 500-511
MDE-AgriVLN: Agricultural Vision-and-Language Navigation with Monocular Depth Estimation
- 19 July 2026
- Published by Springer Nature
- p. 369-381
TomatoScanner: Phenotyping Tomato Fruits Through Only RGB Images
- 19 July 2026
- Published by Springer Nature
- p. 369-381
TomatoScanner: Phenotyping Tomato Fruits Through Only RGB Images
- Open Access
- 8 June 2026
- Published by Elsevier in SSRN Electronic Journal
VPLHOI: Vision, Pose, and Language Fusion for Human-Object Interaction Detection
Human-Object Interaction detection(HOID) has attracted increasing attention in recent years. One-stage methods have achieved strong performance in human-object association. However, two-stage methods often leverage additional information, such as human pose cues and more generalized object representation mappings, which have not yet been fully incorporated into one-stage frameworks. In this paper, we propose an end-to-end architecture that learns complementary knowledge via a multi-teacher distillation strategy to enhance interaction understanding. To effectively align pose and visual information, we explore how different pose estimation paradigms, including top-down and query-based approaches, can be integrated into the HOID framework. Furthermore, a Visual-Language Fusion module is introduced to guide the semantic space of object recognition while incorporating pose information. Experimental results show that our method, VPLHOI, outperforms state-of-the-art HOID approaches on the HICO-DET and V-COCO benchmarks. Extensive ablation studies further demonstrate the effectiveness of the proposed information fusion strategy.
- 8 June 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessVPLHOI: Vision, Pose, and Language Fusion for Human-Object Interaction Detection
Human-Object Interaction detection(HOID) has attracted increasing attention in recent years. One-stage methods have achieved strong performance in human-object association. However, two-stage methods often leverage additional information, such as human pose cues and more generalized object representation mappings, which have not yet been fully incorporated into one-stage frameworks. In this paper, we propose an end-to-end architecture that learns complementary knowledge via a multi-teacher distillation strategy to enhance interaction understanding. To effectively align pose and visual information, we explore how different pose estimation paradigms, including top-down and query-based approaches, can be integrated into the HOID framework. Furthermore, a Visual-Language Fusion module is introduced to guide the semantic space of object recognition while incorporating pose information. Experimental results show that our method, VPLHOI, outperforms state-of-the-art HOID approaches on the HICO-DET and V-COCO benchmarks. Extensive ablation studies further demonstrate the effectiveness of the proposed information fusion strategy.
- 30 March 2026
- Published by Springer Nature in Water Conservation Science and Engineering
- Vol. 11 (1) , 44
A Bayesian Reliability Framework for Improving Irrigation Efficiency in Greenhouse Strawberry Systems
- 30 March 2026
- Published by Springer Nature in Water Conservation Science and Engineering
- Vol. 11 (1) , 44
A Bayesian Reliability Framework for Improving Irrigation Efficiency in Greenhouse Strawberry Systems
- 15 December 2025
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 23 (1) , 29
YOLO11-SPE: a lightweight object detection model for corn seedling counting
- 15 December 2025
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 23 (1) , 29
YOLO11-SPE: a lightweight object detection model for corn seedling counting
- 17 October 2025
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 37 (25-26)
A Study on Malignant Cell Recognition in Rapid On‐Site Cytological Evaluation
Rapid on‐site evaluation (ROSE) technique plays an important role in clinical pathology diagnosis, which can significantly improve the early detection rate and diagnostic efficiency of malignant cells. Facing the ROSE scene, we propose a one‐stage target detection model based on ViT and a lightweight backbone (ShuffleNetV2, MobileNetV3). PyQt5 framework and Flask framework are used to construct the software system and realize the online recognition function. The experimental results show that the model performs well under the mAP and FLOPs metrics, has lower time complexity, and higher target detection accuracy than the traditional model, and provides technical support for ROSE.
- 17 October 2025
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 37 (25-26)
A Study on Malignant Cell Recognition in Rapid On‐Site Cytological Evaluation
Rapid on‐site evaluation (ROSE) technique plays an important role in clinical pathology diagnosis, which can significantly improve the early detection rate and diagnostic efficiency of malignant cells. Facing the ROSE scene, we propose a one‐stage target detection model based on ViT and a lightweight backbone (ShuffleNetV2, MobileNetV3). PyQt5 framework and Flask framework are used to construct the software system and realize the online recognition function. The experimental results show that the model performs well under the mAP and FLOPs metrics, has lower time complexity, and higher target detection accuracy than the traditional model, and provides technical support for ROSE.