iffDetector: Inference-aware Feature Filtering for Object Detection

TL;DR

The paper introduces iffDetector, a novel inference-aware feature filtering module that enhances object detection by optimizing features during both training and inference, leading to improved accuracy with minimal computational overhead.

Contribution

It proposes a generic, feedback-based feature filtering module that can be integrated with existing detectors, improving feature quality during inference.

Findings

Outperforms state-of-the-art methods on PASCAL VOC and MS COCO datasets.

Demonstrates theoretical stability of feature learning via Fourier analysis.

Achieves significant accuracy improvements with negligible computational cost.

Abstract

Modern CNN-based object detectors focus on feature configuration during training but often ignore feature optimization during inference. In this paper, we propose a new feature optimization approach to enhance features and suppress background noise in both the training and inference stages. We introduce a generic Inference-aware Feature Filtering (IFF) module that can easily be combined with modern detectors, resulting in our iffDetector. Unlike conventional open-loop feature calculation approaches without feedback, the IFF module performs closed-loop optimization by leveraging high-level semantics to enhance the convolutional features. By applying Fourier transform analysis, we demonstrate that the IFF module acts as a negative feedback that theoretically guarantees the stability of feature learning. IFF can be fused with CNN-based object detectors in a plug-and-play manner with negligible computational cost overhead. Experiments on the PASCAL VOC and MS COCO datasets demonstrate that our iffDetector consistently outperforms state-of-the-art methods by significant margins\footnote{The test code and model are anonymously available in https://github.com/anonymous2020new/iffDetector }.