Learning a Practical SDR-to-HDRTV Up-conversion using New Dataset and Degradation Models
This work addresses a practical media industry need for SDR-to-HDRTV up-conversion, but it is incremental as it builds on existing learning-based approaches by focusing on dataset and model improvements.
The paper tackled the problem of converting standard dynamic range (SDR) video to high dynamic range (HDRTV) for improved viewing on HDR-WCG TVs, where existing methods produce dim and desaturated results, by proposing a new dataset and degradation models, resulting in enhanced performance validated through tailored metrics and subjective experiments.
In media industry, the demand of SDR-to-HDRTV up-conversion arises when users possess HDR-WCG (high dynamic range-wide color gamut) TVs while most off-the-shelf footage is still in SDR (standard dynamic range). The research community has started tackling this low-level vision task by learning-based approaches. When applied to real SDR, yet, current methods tend to produce dim and desaturated result, making nearly no improvement on viewing experience. Different from other network-oriented methods, we attribute such deficiency to training set (HDR-SDR pair). Consequently, we propose new HDRTV dataset (dubbed HDRTV4K) and new HDR-to-SDR degradation models. Then, it's used to train a luminance-segmented network (LSN) consisting of a global mapping trunk, and two Transformer branches on bright and dark luminance range. We also update assessment criteria by tailored metrics and subjective experiment. Finally, ablation studies are conducted to prove the effectiveness. Our work is available at: https://github.com/AndreGuo/HDRTVDM.