Class Correlation affects Single Object Localization using Pre-trained ConvNets
This addresses the problem of object localization for computer vision researchers by showing how pre-trained models generalize, but it is incremental as it builds on existing ConvNet capabilities.
The paper investigates whether pre-trained ConvNets for classification can localize single objects in images from separate datasets without fine-tuning, using a simple cropping and blackening method called EISS, and finds that class correlation in training data affects localization by making networks respond more to object features than descriptors.
The problem of object localization has become one of the mainstream problems of vision. Most of the algorithms proposed involve the design for the model to be specifically for localizing objects. In this paper, we explore whether a pre-trained canonical ConvNet (without fine-tuning) trained purely for object classification on one dataset with global image level labels can be used to localize objects in images containing a single instance on a separate dataset while generalizing to novel classes. We propose a simple algorithm involving cropping and blackening out regions in the image space called Explicit Image Space based Search (EISS) for locating the most responsive regions in an image in the context of object localization. EISS brings to light the interesting phenomenon of a ConvNets responding more to features within objects as opposed to object level descriptors, as the classes in the training data get more correlated (visually/semantically similar).