RODEO: Replay for Online Object Detection

Abstract

Humans can incrementally learn to do new visual detection tasks, which is a huge challenge for today's computer vision systems. Incrementally trained deep learning models lack backwards transfer to previously seen classes and suffer from a phenomenon known as $"catastrophicforgetting."$ In this paper, we pioneer online streaming learning for object detection, where an agent must learn examples one at a time with severe memory and computational constraints. In object detection, a system must output all bounding boxes for an image with the correct label. Unlike earlier work, the system described in this paper can learn this task in an online manner with new classes being introduced over time. We achieve this capability by using a novel memory replay mechanism that efficiently replays entire scenes. We achieve state-of-the-art results on both the PASCAL VOC 2007 and MS COCO datasets.

Accepted for poster presentation at BMVC2020

References in corpus (6)

• MobileNetV2: Inverted Residuals and Linear Bottlenecks
• Riemannian Walk for Incremental Learning: Understanding Forgetting and Intransigence
• Continual Learning Through Synaptic Intelligence
• FearNet: Brain-Inspired Model for Incremental Learning
• Variational Continual Learning
• Incremental Deep Learning for Robust Object Detection in Unknown Cluttered Environments

Cited by in corpus (5)

• Multi-Task Incremental Learning for Object Detection
• Self-Supervised Training Enhances Online Continual Learning
• Replay in Deep Learning: Current Approaches and Missing Biological Elements
• Towards Open World Object Detection
• Selective Replay Enhances Learning in Online Continual Analogical Reasoning