Mikołaj Jastrzębski · Research

DART: A Degradation-Aware Recurrent Transformer for Archival Film Restoration

Mikołaj Jastrzębski, Wojciech Kozłowski, Kamil Adamczewski
Wrocław University of Science and Technology

Asian Conference on Computer Vision (ACCV) 2026

Paper (arXiv:2607.21219) · PDF · Code

Code and pretrained weights will be released by the end of October 2026.

Abstract

Archival film restoration is a challenging problem because historical footage contains compound degradations such as scratches, dust, blur, noise, flicker, and photometric aging, while clean reference videos are unavailable. Existing video restoration methods largely treat these degradations implicitly, reconstructing frames without explicit knowledge of where damage occurs or how severe it is. We propose DART, a degradation-aware recurrent transformer for archival film restoration. DART predicts and propagates a soft defect mask through time, using it to guide temporal fusion and condition the restoration network on both damage location and severity. This makes the restoration process explicitly aware of film artifacts rather than relying only on reconstruction losses. Experiments on real archival benchmarks show that DART improves no-reference perceptual quality over prior restoration architectures while remaining compact and efficient, producing cleaner and more temporally consistent restorations of structured film damage.

Key results

• State-of-the-art perceptual quality on real archival benchmarks, outperforming BasicVSR++, RTN, DeepRemaster & MambaOFR.
• Compact and efficient: 6.6M parameters at 0.35 GB - among the smallest and most memory-efficient models evaluated.
• Novel multi-scale Dilation Pyramid MaskNet with direct mask supervision and AdaLN-Zero degradation conditioning.

Cite

@inproceedings{jastrzebski2026dart,
  title     = {{DART}: A Degradation-Aware Recurrent Transformer for Archival Film Restoration},
  author    = {Jastrz{\k{e}}bski, Miko{\l}aj and Koz{\l}owski, Wojciech and Adamczewski, Kamil},
  booktitle = {Asian Conference on Computer Vision (ACCV)},
  year      = {2026}
}

Companion paper

AbsoluteDegradation: A Physics-Inspired Synthetic Film-Degradation Pipeline and Archival Film Restoration Benchmark (NeurIPS 2026)