# A Mamba-based Perceptual Loss Function for Learning-based UGC Transcoding

arXiv: https://arxiv.org/abs/2603.25566 · HTML: https://mlsanity.com/paper/2603.25566/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Vision Mamba backbones, one of 10 directions in State-space sequence models (S4, Mamba).
- Builds on 26 of its 46 references that are in the map.
- First posted on arXiv 26 Mar 2026.

## Placed in

- State-space sequence models (S4, Mamba) / [Vision Mamba backbones](https://mlsanity.com/ssm/vision-mamba-backbones/)

## Builds on (most cited first, 26 of 26)

- [The Unreasonable Effectiveness of Deep Features as a Perceptual Metric](https://mlsanity.com/paper/1801.03924/) (2018)
- [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://mlsanity.com/paper/2312.00752/) (2023)
- [Video Enhancement with Task-Oriented Flow](https://mlsanity.com/paper/1711.09078/) (2017)
- [DVC: An End-to-end Deep Video Compression Framework](https://mlsanity.com/paper/1812.00101/) (2018)
- [VideoMamba: State Space Model for Efficient Video Understanding](https://mlsanity.com/paper/2403.06977/) (2024)
- [NeRV: Neural Representations for Videos](https://mlsanity.com/paper/2110.13903/) (2021)
- [Deep Contextual Video Compression](https://mlsanity.com/paper/2109.15047/) (2021)
- [Large-Scale Study of Perceptual Video Quality](https://mlsanity.com/paper/1803.01761/) (2018)
- [YouTube UGC Dataset for Video Compression Research](https://mlsanity.com/paper/1904.06457/) (2019)
- [HiNeRV: Video Compression with Hierarchical Encoding-based Neural Representation](https://mlsanity.com/paper/2306.09818/) (2023)
- [BVI-DVC: A Training Database for Deep Video Compression](https://mlsanity.com/paper/2003.13552/) (2020)
- [Multi-level Wavelet-based Generative Adversarial Network for Perceptual Quality Enhancement of Compressed Video](https://mlsanity.com/paper/2008.00499/) (2020)
- [C3DVQA: Full-Reference Video Quality Assessment with 3D Convolutional Neural Network](https://mlsanity.com/paper/1910.13646/) (2019)
- [Predicting the Quality of Compressed Videos with Pre-Existing Distortions](https://mlsanity.com/paper/2004.02943/) (2020)
- [UGC-VIDEO: perceptual quality assessment of user-generated videos](https://mlsanity.com/paper/1908.11517/) (2019)
- [RankDVQA: Deep VQA based on Ranking-inspired Hybrid Training](https://mlsanity.com/paper/2202.08595/) (2022)
- [DVC-P: Deep Video Compression with Perceptual Optimizations](https://mlsanity.com/paper/2109.10849/) (2021)
- [Good, Cheap, and Fast: Overfitted Image Compression with Wasserstein Distortion](https://mlsanity.com/paper/2412.00505/) (2024)
- [CVEGAN: A Perceptually-inspired GAN for Compressed Video Enhancement](https://mlsanity.com/paper/2011.09190/) (2020)
- [GIViC: Generative Implicit Video Compression](https://mlsanity.com/paper/2503.19604/) (2025)
- [Adversarial Distortion for Learned Video Compression](https://mlsanity.com/paper/2004.09508/) (2020)
- [Video Compression with CNN-based Post Processing](https://mlsanity.com/paper/2009.07583/) (2020)
- [Enhancing VMAF through New Feature Integration and Model Combination](https://mlsanity.com/paper/2103.06338/) (2021)
- [Full-reference Video Quality Assessment for User Generated Content Transcoding](https://mlsanity.com/paper/2312.12317/) (2023)
- [Rate-Distortion Optimization With Alternative References For UGC Video Compression](https://mlsanity.com/paper/2303.06254/) (2023)
- [BVI-UGC: A Video Quality Database for User-Generated Content Transcoding](https://mlsanity.com/paper/2408.07171/) (2024)

## Abstract (arXiv)

In user-generated content (UGC) transcoding, source videos typically suffer various degradations due to prior compression, editing, or suboptimal capture conditions. Consequently, existing video compression paradigms that solely optimize for fidelity relative to the reference become suboptimal, as they force the codec to replicate the inherent artifacts of the non-pristine source. To address this, we propose a novel perceptually inspired loss function for learning-based UGC video transcoding that redefines the role of the reference video, shifting it from a ground-truth pixel anchor to an informative contextual guide. Specifically, we train a lightweight neural quality model based on a Selective Structured State-Space Model (Mamba) optimized using a weakly-supervised Siamese ranking strategy. The proposed model is then integrated into the rate-distortion optimization (RDO) process of two neural video codecs (DCVC and HiNeRV) as a loss function, aiming to generate reconstructed content with improved perceptual quality. Our experiments demonstrate that this framework achieves substantial coding gains over both autoencoder and implicit neural representation-based baselines, with 8.46% and 12.89% BD-rate savings, respectively.