Artificial Intelligence Medical Compendium

Diffusion models achieve high-quality image generation but remain computationally expensive due to iterative denoising across many timesteps. Existing acceleration methods mainly reduce per-sample latency, but gains along this axis are becoming satur... read more 

Existing deepfake detection models are built upon manipu- lated images that are tampered with either exclusively fully swapped or fully synthesised faces, which leads us to an open question: Can off-the-shelf deepfake models detect manipulation effec... read more 

Unified multimodal models (UMMs) receive image-generation supervision that exposes them to visual information not fully documented in text. Can learning to generate images expand what pretrained models know? We introduce Knowledge Transfer Probing (K... read more 

We study Text-and-Image-to-Image (TI2I) flow matching methods for controllable image editing. In a broader generative modeling context, Classifier-Free Guidance (CFG) is established as a key component responsible for high quality outputs. However, we... read more 

Diffusion models achieve strong performance in text-to-image generation, but they can still produce unsafe content. Training-free safety guidance offers a practical inference-time alternative that intervenes during sampling without modifying model we... read more 

Decision trees can make vision-language model (VLM) classification interpretable by replacing a single prediction with a sequence of root-to-leaf questions. However, the accuracy cost of this greedy, non-backtracking procedure remains unclear. We int... read more 

Recent models generate sewing patterns from images or text, and program-based representations make these patterns executable. These systems produce a finished pattern and do not model how a patternmaker turns one pattern into another. In practice a n... read more 

Medical segmentation uncertainty is often evaluated with calibration error or pixel-level error-detection scores, but these metrics do not show whether uncertainty improves downstream review decisions. We propose a decision-aware evaluation protocol ... read more 

Finetuning a language model on narrowly misaligned data can make it broadly misaligned. Interleaving generic safety data is the most effective known in-training defense, but the misalignment returns as soon as a prompt resembles the finetuning contex... read more 

Recent feed-forward 3D Gaussian Splatting (3DGS) methods have enabled effective multi-view 3D reconstruction by predicting Gaussian primitives in a single forward pass. However, existing approaches predominantly adopt a pixel-aligned paradigm that pr... read more