Search papers, labs, and topics across Lattice.
This paper introduces DART, a degradation-aware recurrent transformer designed specifically for the restoration of archival films suffering from various compound degradations. By predicting and propagating a soft defect mask through time, DART effectively guides the restoration process, explicitly accounting for damage location and severity rather than relying solely on reconstruction losses. Experimental results on real archival benchmarks demonstrate that DART significantly enhances no-reference perceptual quality compared to existing methods, resulting in cleaner and more temporally consistent restorations of damaged films.
DART achieves superior archival film restoration by explicitly modeling degradation, leading to cleaner and more temporally consistent outputs than previous methods.
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.
