GAIA: A Transfer Learning System of Object Detection that Fits Your Needs

TL;DR

GAIA is a transfer learning system for object detection that efficiently produces customized models tailored to specific latency, data, and resource constraints, demonstrating strong results across multiple datasets.

Contribution

GAIA introduces an automated system that generates tailored object detection models for diverse downstream needs, addressing the challenge of large-scale pre-training costs.

Findings

Achieves AP from 38.2 to 46.5 on COCO

Produces models with latency from 16ms to 53ms

Effective across multiple datasets including COCO, Objects365, and others

Abstract

Transfer learning with pre-training on large-scale datasets has played an increasingly significant role in computer vision and natural language processing recently. However, as there exist numerous application scenarios that have distinctive demands such as certain latency constraints and specialized data distributions, it is prohibitively expensive to take advantage of large-scale pre-training for per-task requirements. In this paper, we focus on the area of object detection and present a transfer learning system named GAIA, which could automatically and efficiently give birth to customized solutions according to heterogeneous downstream needs. GAIA is capable of providing powerful pre-trained weights, selecting models that conform to downstream demands such as latency constraints and specified data domains, and collecting relevant data for practitioners who have very few datapoints for their tasks. With GAIA, we achieve promising results on COCO, Objects365, Open Images, Caltech, CityPersons, and UODB which is a collection of datasets including KITTI, VOC, WiderFace, DOTA, Clipart, Comic, and more. Taking COCO as an example, GAIA is able to efficiently produce models covering a wide range of latency from 16ms to 53ms, and yields AP from 38.2 to 46.5 without whistles and bells. To benefit every practitioner in the community of object detection, GAIA is released at https://github.com/GAIA-vision.