A Random CNN Sees Objects: One Inductive Bias of CNN and Its Applications
This addresses the problem of improving self-supervised learning for computer vision, particularly object detection, by leveraging an inherent CNN bias, offering a novel but incremental approach.
The paper reveals that randomly initialized CNNs can localize objects without training, an inductive bias named Tobias, and applies it to self-supervised learning by guiding foreground-background separation to improve object detection and other tasks with consistent gains across dataset sizes and augmentations.
This paper starts by revealing a surprising finding: without any learning, a randomly initialized CNN can localize objects surprisingly well. That is, a CNN has an inductive bias to naturally focus on objects, named as Tobias ("The object is at sight") in this paper. This empirical inductive bias is further analyzed and successfully applied to self-supervised learning (SSL). A CNN is encouraged to learn representations that focus on the foreground object, by transforming every image into various versions with different backgrounds, where the foreground and background separation is guided by Tobias. Experimental results show that the proposed Tobias significantly improves downstream tasks, especially for object detection. This paper also shows that Tobias has consistent improvements on training sets of different sizes, and is more resilient to changes in image augmentations. Code is available at https://github.com/CupidJay/Tobias.