Differentiable Product Quantization for Memory Efficient Camera Relocalization

TL;DR

This paper introduces a differentiable auto-encoder approach for descriptor quantization that balances memory efficiency and relocalization accuracy, outperforming existing methods on Aachen Day-Night.

Contribution

It proposes a novel end-to-end trainable auto-encoder for descriptor compression that preserves matching performance, enabling highly memory-efficient camera relocalization.

Findings

Achieves superior relocalization accuracy with only 1MB descriptor memory.

Outperforms existing compression methods on Aachen Day-Night dataset.

Demonstrates effective trade-off between memory reduction and localization performance.

Abstract

Camera relocalization relies on 3D models of the scene with a large memory footprint that is incompatible with the memory budget of several applications. One solution to reduce the scene memory size is map compression by removing certain 3D points and descriptor quantization. This achieves high compression but leads to performance drop due to information loss. To address the memory performance trade-off, we train a light-weight scene-specific auto-encoder network that performs descriptor quantization-dequantization in an end-to-end differentiable manner updating both product quantization centroids and network parameters through back-propagation. In addition to optimizing the network for descriptor reconstruction, we encourage it to preserve the descriptor-matching performance with margin-based metric loss functions. Results show that for a local descriptor memory of only 1MB, the synergistic combination of the proposed network and map compression achieves the best performance on the Aachen Day-Night compared to existing compression methods.