INTERLACE removes 25% of a vision-language model's layers while keeping about 89% of its performance in challenging benchmarks.
The presence of a foveal gazer meaningfully improves accuracy in locating a gaze goal far in the periphery
Analyizing the inner workings of CNNs to predict neuron types mediating covert attention
CNNs discount target-like structures of normal anatomic structures like radiologists do.
It might help explain attention-like behaviors in simple organisms from mice to fruit flies.
When and how AI helps radiologists. Answers from vision science.
Eye lookers and mouth lookers' brains process faces differently
A new paper in ElIfe by recent lab Ph.D. Nicole Han characterizes the inferential and functional nature of social attention’s fine-grain eye movement dynamics.
Why do radiologists rely on 2D synthesized images to complement their search through 3D image stacks?
Her dissertation advances the understanding of gaze perception
