LLC: Accurate, Multi-purpose Learnt Low-dimensional Binary Codes

TL;DR

This paper introduces LLC, a novel method for learning extremely low-dimensional binary codes for images that achieve high accuracy in classification, retrieval, and out-of-distribution detection without requiring side information.

Contribution

LLC is a new approach that learns low-dimensional binary representations for instances and classes without side-information, achieving near-optimal accuracy with very compact codes (~20 bits).

Findings

Outperforms 16-bit HashNet with only 10 bits on ImageNet-100 retrieval.

Achieves classification accuracy comparable to real-valued features.

Enables out-of-distribution detection without threshold tuning.

Abstract

Learning binary representations of instances and classes is a classical problem with several high potential applications. In modern settings, the compression of high-dimensional neural representations to low-dimensional binary codes is a challenging task and often require large bit-codes to be accurate. In this work, we propose a novel method for Learning Low-dimensional binary Codes (LLC) for instances as well as classes. Our method does not require any side-information, like annotated attributes or label meta-data, and learns extremely low-dimensional binary codes (~20 bits for ImageNet-1K). The learnt codes are super-efficient while still ensuring nearly optimal classification accuracy for ResNet50 on ImageNet-1K. We demonstrate that the learnt codes capture intrinsically important features in the data, by discovering an intuitive taxonomy over classes. We further quantitatively measure the quality of our codes by applying it to the efficient image retrieval as well as out-of-distribution (OOD) detection problems. For ImageNet-100 retrieval problem, our learnt binary codes outperform 16 bit HashNet using only 10 bits and also are as accurate as 10 dimensional real representations. Finally, our learnt binary codes can perform OOD detection, out-of-the-box, as accurately as a baseline that needs ~3000 samples to tune its threshold, while we require none. Code is open-sourced at https://github.com/RAIVNLab/LLC.