Point-Level Region Contrast for Object Detection Pre-Training

TL;DR

This paper introduces point-level region contrast, a self-supervised pre-training method that enhances object detection by balancing localization and recognition through contrastive learning on individual points.

Contribution

It proposes a novel point-level contrastive pre-training approach that improves robustness to region quality and enables online knowledge distillation for better object detection.

Findings

Outperforms state-of-the-art pre-training methods on detection and segmentation tasks.

Improves robustness to imperfect region proposals during training.

Provides extensive ablation studies and visualizations.

Abstract

In this work we present point-level region contrast, a self-supervised pre-training approach for the task of object detection. This approach is motivated by the two key factors in detection: localization and recognition. While accurate localization favors models that operate at the pixel- or point-level, correct recognition typically relies on a more holistic, region-level view of objects. Incorporating this perspective in pre-training, our approach performs contrastive learning by directly sampling individual point pairs from different regions. Compared to an aggregated representation per region, our approach is more robust to the change in input region quality, and further enables us to implicitly improve initial region assignments via online knowledge distillation during training. Both advantages are important when dealing with imperfect regions encountered in the unsupervised setting. Experiments show point-level region contrast improves on state-of-the-art pre-training methods for object detection and segmentation across multiple tasks and datasets, and we provide extensive ablation studies and visualizations to aid understanding. Code will be made available.