Few-shot Weakly-Supervised Object Detection via Directional Statistics

TL;DR

This paper introduces a probabilistic approach using von Mises-Fisher distributions for few-shot weakly-supervised object detection, leveraging pre-trained features and EM steps to improve localization and detection of novel objects with minimal supervision.

Contribution

It proposes a novel probabilistic multiple instance learning method employing vMF distributions for few-shot object detection using only image-level labels, outperforming existing methods.

Findings

Outperforms strong baselines in few-shot COL and WSOD

Effective use of pre-trained features from Faster-RCNN

Simple yet powerful probabilistic modeling approach

Abstract

Detecting novel objects from few examples has become an emerging topic in computer vision recently. However, these methods need fully annotated training images to learn new object categories which limits their applicability in real world scenarios such as field robotics. In this work, we propose a probabilistic multiple instance learning approach for few-shot Common Object Localization (COL) and few-shot Weakly Supervised Object Detection (WSOD). In these tasks, only image-level labels, which are much cheaper to acquire, are available. We find that operating on features extracted from the last layer of a pre-trained Faster-RCNN is more effective compared to previous episodic learning based few-shot COL methods. Our model simultaneously learns the distribution of the novel objects and localizes them via expectation-maximization steps. As a probabilistic model, we employ von Mises-Fisher (vMF) distribution which captures the semantic information better than Gaussian distribution when applied to the pre-trained embedding space. When the novel objects are localized, we utilize them to learn a linear appearance model to detect novel classes in new images. Our extensive experiments show that the proposed method, despite being simple, outperforms strong baselines in few-shot COL and WSOD, as well as large-scale WSOD tasks.