• Weak and Small Object Detection and Recognition Algorithms (弱小目标检测与识别算法)
• Object Perception under Clutter, Low Contrast, and Occlusion (复杂背景、低对比度与遮挡条件下的目标感知)
• Multi-scale Feature Learning, Super-resolution, and Fine-grained Vision (多尺度特征学习、超分辨率与细粒度视觉)
  
• Deep Learning, Vision Transformers, and Foundation Models for Weak and Small Objects (基于深度学习、视觉Transformer及基础模型的弱小目标识别)
• Multimodal Fusion of Infrared, Visible, Radar, and Hyperspectral Data (红外、可见光、雷达与高光谱多模态融合)
• Weak and Small Object Analysis in Remote Sensing, UAV, and Satellite Video (遥感影像、无人机航拍与卫星视频中的弱小目标分析)
• Weak and Small Object Tracking, Spatiotemporal Modeling, and Video Understanding (弱小目标跟踪、时空建模与视频目标理解)
• Few-shot, Zero-shot, Weakly Supervised, and Self-supervised Learning (小样本、零样本、弱监督与自监督学习)
• Dataset Construction, Data Augmentation, Benchmarks, and Trustworthy Evaluation (数据集构建、数据增强、评测基准与可信评估)
• Lightweight Models and Real-time Deployment on Embedded and Edge Devices (面向嵌入式设备与边缘端的轻量化、实时部署)

UWeak and small object recognition remains a critical challenge in computer vision and intelligent sensing due to limited pixel representation, cluttered backgrounds, large scale variations, occlusions, and adverse imaging conditions. This track targets researchers and practitioners in vision, image processing, remote sensing, autonomous systems, industrial inspection, and edge intelligence. It focuses on detection, recognition, tracking, and multimodal fusion for weak and small objects. Participants will gain insights into state-of-the-art algorithms, datasets, benchmarks, and real-world deployment practices, while discussing practical solutions that jointly achieve high accuracy, robustness, and low computational cost.