The Flow has not summarised this story yet — read it at arXiv Machine Learning.
The paper introduces a dual‑stage deep learning system for detecting dental caries in panoramic radiographs. It first localizes teeth using Faster R‑CNN, then applies U‑Net for pixel‑wise caries segmentation, converting polygon annotations into high‑resolution binary masks. Trained on 3,000 images with both expert and algorithmic labels, the model achieves an IoU of 0.9013, Dice of 0.9482, Recall of 0.9433, and Precision of 0.9774, outperforming existing methods and reducing false positives.
arXiv:2610.15061v1 Announce Type: cross Abstract: Accurate and efficient analysis of dental radiographs has become increasingly important for disease diagnosis and clinical decision making. In partic...
arXiv:2607. 22139v1 Announce Type: cross Abstract: Accurate pixel-level classification of coronary angiograms is critical for cardiovascular disease assessment, yet the field lacks standardized evaluation protocols.
The paper studies Graph-Guided Token Merging (G2TM), a module that reduces token count in Vision Transformers. It evaluates G2TM across multiple segmentation frameworks and decoder types, finding that its performance gains are tied to the encoder rather than the decoder. The authors report consistent reductions in GFLOPs (22‑47%) and throughput improvements (up to 74%) on ADE20K, with optimal hyperparameters depending mainly on backbone pre‑training and target dataset.
arXiv:2410.07421v2 Announce Type: replace Abstract: Instance segmentation is a core computer vision task with great practical significance. Recent advances, driven by large-scale benchmark datasets,...
The paper introduces D3S2, a diffusion‑guided dataset distillation framework tailored for semantic segmentation. It tackles long‑tailed class imbalance, pixel‑wise alignment, and high computational cost by first selecting class‑balanced masks and then synthesizing images with a pretrained diffusion model conditioned on those masks. Guided diffusion sampling further refines the data with segmentation‑consistency and class‑wise feature matching losses, achieving superior performance at a 1% compression rate on ADE20K and COCO‑Stuff.