Redundant Queries in DETR-Based 3D Detection Methods: Unnecessary and Prunable

TL;DR

This paper introduces GPQ, a simple query pruning method for DETR-based 3D detection models that reduces redundant queries, accelerates inference, and decreases computational costs without sacrificing detection performance.

Contribution

The paper proposes a parameter-free, easy-to-implement query pruning technique called GPQ that can be applied as a fine-tuning step to improve efficiency of existing 3D detectors.

Findings

GPQ reduces redundant queries effectively.

Inference speed improves by up to 1.35x on desktops.

FLOPs and inference time decrease significantly on edge devices.

Abstract

Query-based models are extensively used in 3D object detection tasks, with a wide range of pre-trained checkpoints readily available online. However, despite their popularity, these models often require an excessive number of object queries, far surpassing the actual number of objects to detect. The redundant queries result in unnecessary computational and memory costs. In this paper, we find that not all queries contribute equally -- a significant portion of queries have a much smaller impact compared to others. Based on this observation, we propose an embarrassingly simple approach called Gradually Pruning Queries (GPQ), which prunes queries incrementally based on their classification scores. A key advantage of GPQ is that it requires no additional learnable parameters. It is straightforward to implement in any query-based method, as it can be seamlessly integrated as a fine-tuning step using an existing checkpoint after training. With GPQ, users can easily generate multiple models with fewer queries, starting from a checkpoint with an excessive number of queries. Experiments on various advanced 3D detectors show that GPQ effectively reduces redundant queries while maintaining performance. Using our method, model inference on desktop GPUs can be accelerated by up to 1.35x. Moreover, after deployment on edge devices, it achieves up to a 67.86% reduction in FLOPs and a 65.16% decrease in inference time. The code will be available at https://github.com/iseri27/Gpq.