AI Mental Health Training Risks: Clinical Harm Revealed
New research shows AI safety training in mental health can cause harm. Rigorous multi-axis testing is crucial for safe therapeutic AI deployment.
Knee-xRAI: Explainable AI for Accurate Knee Osteoarthritis Grading
Discover Knee-xRAI, an explainable AI framework that improves accuracy and transparency in Kellgren-Lawrence grading of knee osteoarthritis.
Sphere-Depth Benchmark for Robust Spherical Depth Estimation
Sphere-Depth benchmark evaluates depth estimation models under varying spherical camera poses for improved 360° vision and robotic navigation.
Automating Scientific Text Categorization with LLMs & Prompt Chaining
Explore how Large Language Models using prompt chaining improve scientific text categorization for efficient knowledge discovery and research navigation.
PushupBench Reveals VLMs Fail to Count Pushups Accurately
PushupBench dataset exposes large vision-language models' struggle to count pushup reps, highlighting the need for better temporal reasoning in AI video an...