Edge-case Synthesis for Fisheye Object Detection: A Data-centric Perspective

TL;DR

This paper presents a data-centric approach to improve fisheye object detection by synthesizing edge-case images that target model blind spots, leading to enhanced detection performance.

Contribution

It introduces a systematic pipeline for identifying and synthesizing critical edge-cases in fisheye images to address model weaknesses.

Findings

Performance gains from synthetic edge-case data

Effective identification of model blind spots

Improved detection accuracy in challenging scenarios

Abstract

Fisheye cameras introduce significant distortion and pose unique challenges to object detection models trained on conventional datasets. In this work, we propose a data-centric pipeline that systematically improves detection performance by focusing on the key question of identifying the blind spots of the model. Through detailed error analysis, we identify critical edge-cases such as confusing class pairs, peripheral distortions, and underrepresented contexts. Then we directly address them through edge-case synthesis. We fine-tuned an image generative model and guided it with carefully crafted prompts to produce images that replicate real-world failure modes. These synthetic images are pseudo-labeled using a high-quality detector and integrated into training. Our approach results in consistent performance gains, highlighting how deeply understanding data and selectively fixing its weaknesses can be impactful in specialized domains like fisheye object detection.