Timezone: »

 
Poster
Do Convnets Learn Correspondence?
Jonathan L Long · Ning Zhang · Trevor Darrell

Tue Dec 09 04:00 PM -- 08:59 PM (PST) @ Level 2, room 210D
in Poster Session »

Convolutional neural nets (convnets) trained from massive labeled datasets have substantially improved the state-of-the-art in image classification and object detection. However, visual understanding requires establishing correspondence on a finer level than object category. Given their large pooling regions and training from whole-image labels, it is not clear that convnets derive their success from an accurate correspondence model which could be used for precise localization. In this paper, we study the effectiveness of convnet activation features for tasks requiring correspondence. We present evidence that convnet features localize at a much finer scale than their receptive field sizes, that they can be used to perform intraclass aligment as well as conventional hand-engineered features, and that they outperform conventional features in keypoint prediction on objects from PASCAL VOC 2011.

[ PDF] 

Author Information

Jonathan L Long (UC Berkeley)
Ning Zhang (UC Berkeley)
Trevor Darrell (UC Berkeley)

More from the Same Authors