Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

FlashClear: Ultra-Fast Image Content Removal via Efficient Step Distillation and Feature Caching

About

Recently, diffusion-based object removal models have achieved impressive results in eliminating objects and their associated visual effects. However, they indiscriminately denoise all tokens across all timesteps, ignoring that removal usually involves small foreground regions. This strategy introduces substantial computational overhead and prolonged inference times. To overcome this computational burden, we propose a latent discriminator to implement Region-aware Adversarial Distillation (RAD), yielding a highly efficient few-step model named FlashClear. Furthermore, tailored to few-step diffusion models, we propose FPAC (Foreground-Prioritized Asymmetric Attention and Caching), a training-free acceleration strategy. Extensive experiments demonstrate that our framework provides massive acceleration while maintaining or exceeding the performance of our base model, ObjectClear. Notably, on the OBER benchmark, our FlashClear achieves up to 8.26$\times$ and 122$\times$ speedup over ObjectClear and OmniPaint, respectively, while maintaining high visual quality and fidelity.

Yixin Tang, Jiawei Guo, Junxian Li, Zhiteng Li, Jixin Zhao, Bingya Zhang, Chenbo Wang, Yulun Zhang, Shangchen Zhou• 2026

Related benchmarks

TaskDatasetResultRank
Object RemovalOBER 512x512 (test)
LPIPS0.0351
10
Object RemovalRORD 960x540 (val)
LPIPS0.0698
10
Object RemovalUser Study against ObjectClear
Win Rate22.5
2
Showing 3 of 3 rows

Other info

Follow for update