LSDA: Large Scale Detection Through Adaptation

TL;DR

This paper introduces LSDA, a method that leverages large-scale classification data to enable detection for many categories lacking bounding box annotations, significantly expanding detection capabilities.

Contribution

The paper presents a novel adaptation approach that transfers knowledge from classification to detection, enabling large-scale detection without extensive bounding box annotations.

Findings

Successfully trained a detector for over 7,600 categories.

Demonstrated the effectiveness of transfer learning from classification to detection.

Achieved a detection speed of 2 frames per second for large-scale detectors.

Abstract

A major challenge in scaling object detection is the difficulty of obtaining labeled images for large numbers of categories. Recently, deep convolutional neural networks (CNNs) have emerged as clear winners on object classification benchmarks, in part due to training with 1.2M+ labeled classification images. Unfortunately, only a small fraction of those labels are available for the detection task. It is much cheaper and easier to collect large quantities of image-level labels from search engines than it is to collect detection data and label it with precise bounding boxes. In this paper, we propose Large Scale Detection through Adaptation (LSDA), an algorithm which learns the difference between the two tasks and transfers this knowledge to classifiers for categories without bounding box annotated data, turning them into detectors. Our method has the potential to enable detection for the tens of thousands of categories that lack bounding box annotations, yet have plenty of classification data. Evaluation on the ImageNet LSVRC-2013 detection challenge demonstrates the efficacy of our approach. This algorithm enables us to produce a >7.6K detector by using available classification data from leaf nodes in the ImageNet tree. We additionally demonstrate how to modify our architecture to produce a fast detector (running at 2fps for the 7.6K detector). Models and software are available at