Improving Long-tailed Object Detection with Image-Level Supervision by Multi-Task Collaborative Learning

TL;DR

This paper introduces CLIS, a multi-task collaborative framework leveraging image-level supervision to improve long-tailed object detection, especially for tail categories, achieving state-of-the-art results on LVIS dataset.

Contribution

Proposes a novel multi-task collaborative learning framework that effectively utilizes image-level supervision to enhance tail category detection in long-tailed datasets.

Findings

Achieves 31.1 AP on LVIS dataset, surpassing previous methods.

Improves tail category AP by 10.1 points.

Demonstrates effectiveness without complex loss engineering.

Abstract

Data in real-world object detection often exhibits the long-tailed distribution. Existing solutions tackle this problem by mitigating the competition between the head and tail categories. However, due to the scarcity of training samples, tail categories are still unable to learn discriminative representations. Bringing more data into the training may alleviate the problem, but collecting instance-level annotations is an excruciating task. In contrast, image-level annotations are easily accessible but not fully exploited. In this paper, we propose a novel framework CLIS (multi-task Collaborative Learning with Image-level Supervision), which leverage image-level supervision to enhance the detection ability in a multi-task collaborative way. Specifically, there are an object detection task (consisting of an instance-classification task and a localization task) and an image-classification task in our framework, responsible for utilizing the two types of supervision. Different tasks are trained collaboratively by three key designs: (1) task-specialized sub-networks that learn specific representations of different tasks without feature entanglement. (2) a siamese sub-network for the image-classification task that shares its knowledge with the instance-classification task, resulting in feature enrichment of detectors. (3) a contrastive learning regularization that maintains representation consistency, bridging feature gaps of different supervision. Extensive experiments are conducted on the challenging LVIS dataset. Without sophisticated loss engineering, CLIS achieves an overall AP of 31.1 with 10.1 point improvement on tail categories, establishing a new state-of-the-art. Code will be at https://github.com/waveboo/CLIS.