FUSE: Label-Free Image-Event Joint Monocular Depth Estimation via Frequency-Decoupled Alignment and Degradation-Robust Fusion

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Sun, Pihai, Jiang, Junjun, Yao, Yuanqi, Chen, Youyu, Zhao, Wenbo, Jiang, Kui, Liu, Xianming
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866912682020438016
author Sun, Pihai
Jiang, Junjun
Yao, Yuanqi
Chen, Youyu
Zhao, Wenbo
Jiang, Kui
Liu, Xianming
author_facet Sun, Pihai
Jiang, Junjun
Yao, Yuanqi
Chen, Youyu
Zhao, Wenbo
Jiang, Kui
Liu, Xianming
contents Image-event joint depth estimation methods leverage complementary modalities for robust perception, yet face challenges in generalizability stemming from two factors: 1) limited annotated image-event-depth datasets causing insufficient cross-modal supervision, and 2) inherent frequency mismatches between static images and dynamic event streams with distinct spatiotemporal patterns, leading to ineffective feature fusion. To address this dual challenge, we propose Frequency-decoupled Unified Self-supervised Encoder (FUSE) with two synergistic components: The Parameter-efficient Self-supervised Transfer (PST) establishes cross-modal knowledge transfer through latent space alignment with image foundation models, effectively mitigating data scarcity by enabling joint encoding without depth ground truth. Complementing this, we propose the Frequency-Decoupled Fusion module (FreDFuse) to explicitly decouple high-frequency edge features from low-frequency structural components, resolving modality-specific frequency mismatches through physics-aware fusion. This combined approach enables FUSE to construct a universal image-event encoder that only requires lightweight decoder adaptation for target datasets. Extensive experiments demonstrate state-of-the-art performance with 14% and 24.9% improvements in Abs .Rel on MVSEC and DENSE datasets. The framework exhibits remarkable zero-shot adaptability to challenging scenarios including extreme lighting and motion blur, significantly advancing real-world deployment capabilities. The source code for our method is publicly available at: https://github.com/sunpihai-up/FUSE
format Preprint
id arxiv_https___arxiv_org_abs_2503_19739
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle FUSE: Label-Free Image-Event Joint Monocular Depth Estimation via Frequency-Decoupled Alignment and Degradation-Robust Fusion
Sun, Pihai
Jiang, Junjun
Yao, Yuanqi
Chen, Youyu
Zhao, Wenbo
Jiang, Kui
Liu, Xianming
Computer Vision and Pattern Recognition
Image-event joint depth estimation methods leverage complementary modalities for robust perception, yet face challenges in generalizability stemming from two factors: 1) limited annotated image-event-depth datasets causing insufficient cross-modal supervision, and 2) inherent frequency mismatches between static images and dynamic event streams with distinct spatiotemporal patterns, leading to ineffective feature fusion. To address this dual challenge, we propose Frequency-decoupled Unified Self-supervised Encoder (FUSE) with two synergistic components: The Parameter-efficient Self-supervised Transfer (PST) establishes cross-modal knowledge transfer through latent space alignment with image foundation models, effectively mitigating data scarcity by enabling joint encoding without depth ground truth. Complementing this, we propose the Frequency-Decoupled Fusion module (FreDFuse) to explicitly decouple high-frequency edge features from low-frequency structural components, resolving modality-specific frequency mismatches through physics-aware fusion. This combined approach enables FUSE to construct a universal image-event encoder that only requires lightweight decoder adaptation for target datasets. Extensive experiments demonstrate state-of-the-art performance with 14% and 24.9% improvements in Abs .Rel on MVSEC and DENSE datasets. The framework exhibits remarkable zero-shot adaptability to challenging scenarios including extreme lighting and motion blur, significantly advancing real-world deployment capabilities. The source code for our method is publicly available at: https://github.com/sunpihai-up/FUSE
title FUSE: Label-Free Image-Event Joint Monocular Depth Estimation via Frequency-Decoupled Alignment and Degradation-Robust Fusion
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2503.19739