name: “Learning Local-Global Contextual Adaptation for Multi-Person Pose Estimation” description: “Multi-person pose estimation using local-global contextual adaptation.” url: “https://github.com/cherubicXN/logocap” features: – “Local-global context adaptation” – “Multi-person pose detection” – “Robust performance in crowded scenarios” usage: – “Pose estimation in real-time video feeds” – “Analyzing human interactions in social settings” name: “Decoupling Makes Weakly Supervised Local Feature Better” description: “Improving weakly supervised local feature learning through decoupling.” url: “https://github.com/The-Learning-And-Vision-Atelier-LAVA/PoSFeat” features: – “Decoupled training strategy” – “Enhanced feature representation” – “Weakly supervised learning capability” usage: – “Image classification tasks” – “Object detection in weakly labeled datasets” name: “SeqDeepFake: Detecting and Recovering Sequential DeepFake Manipulation” description: “Detecting and recovering from sequential deepfake manipulations.” url: “https://github.com/rshaojimmy/SeqDeepFake” features: – “Sequential deepfake detection” – “Manipulation recovery techniques” – “Robust against various deepfake methods” usage: – “Social media content verification” – “Forensic analysis of video evidence” name: “NoiLIn: Do Noisy Labels Always Hurt Adversarial Training?” description: “Investigating the impact of noisy labels on adversarial training.” url: “https://github.com/zjfheart/NoiLIn” features: – “Analysis of noisy label effects” – “Improved adversarial training techniques” – “Robustness against label noise” usage: – “Training robust machine learning models” – “Evaluating dataset quality in adversarial settings” name: “High-Resolution Virtual Try-On with Misalignment and Occlusion-Handled Conditions” description: “Virtual try-on system that handles misalignment and occlusions effectively.” url: “https://github.com/sangyun884/HR-VITON” features: – “High-resolution image processing” – “Occlusion handling capabilities” – “Realistic virtual try-on experience” usage: – “E-commerce applications for clothing” – “Augmented reality fashion apps” name: “Fast and Robust Non-Rigid Registration Using Accelerated Majorization-Minimization” description: “Non-rigid registration method using accelerated majorization-minimization.” url: “https://github.com/yaoyx689/AMM_NRR” features: – “Fast non-rigid registration” – “Majorization-minimization optimization” – “Robustness to noise and artifacts” usage: – “Medical image alignment tasks” – “3D shape matching and registration”-加速的非刚性配准方法
使用加速的极大极小化方法进行非刚性配准，具有抵抗噪声和伪影的能力。
