This paper reports on the NTIRE 2026 Challenge on Bitstream-Corrupted Video Restoration (BSCVR). The challenge aims to advance research on recovering visually coherent videos from corrupted bitstreams, whose decoding often produces severe spatial-temporal artifacts and content distortion. Built upon recent progress in bitstream-corrupted video recovery, the challenge provides a common benchmark for evaluating restoration methods under realistic corruption settings. We describe the dataset, evaluation protocol, and participating methods, and summarize the final results and main technical trends. The challenge highlights the difficulty of this emerging task and provides useful insights for future research on robust video restoration under practical bitstream corruption.
Figures & tables
Figure 1: Video corruption pattern in bitstream-corrupted video recovery problem summarized by [ 33 ] . Compared with the simulated video corruption in existing inpainting or error concealment research, BSCVR contains various realistic corruption patterns including (1) block artifacts (artfs.), (2) color artifacts, (3) duplication artifacts, (4) misalignment, (5) texture loss, (6) trailing artifacts, which is closer to the corrupted video in real world.
NTIRE 2026 Challenge on Bitstream-Corrupted Video Restoration Results
Team
Codabench User
PSNR ↑ (Primary)
SSIM ↑
LPIPS ↓
Rank PSNR
Rank SSIM
Final Rank
MGTV-AI
nerror
33.642
0.9334
0.0900
1
2
1
RedMediaTech
chenyuxiang
32.865
0.9344
0.0852
2
1
2
bighit
hyena
27.873
0.8933
0.1388
3
3
3
Vroom
priyansh
27.370
0.8713
0.2028
4
5
4
weichow
weichow
27.276
0.8727
0.1866
5
4
5
Table 1: NTIRE 2026 Bitstream-Corrupted Video Restoration Challenge results, final rankings, and the main characteristics of the solutions. Note that, the average PSNR value achieved on the test set is used for final ranking.
Figure 2: Visualization of the final results of participants, along with corresponding input reference frames and GT images. Each column represents a video of a different scene. Among them, (a-i) respectively represent the results of Bitstream Corrupted Input, GT, MGTV-AI team, RedMediaTech team, Bighit team, Vroom team, Weichow team, Holding team, and NTR team.
Figure 3: The MGTV-AI Team: Three-Stage Framework For BitStream-corrupted Video Restoration.
Figure 4: The RedMediaTech team proposed a single-step video restoration framework based on Wan2.1 DiT with a two-stage training strategy. The first stage leverages MSE and LPIPS losses to exploit strong generative priors for perceptual quality, while the second stage fine-tunes with MSE loss to improve distortion metrics (PSNR/SSIM). Additionally, replacing the original VAE with Qwen-Image VAE enhances robustness to large motion and complex temporal variations.
Figure 5: The bighit team proposed two-stage framework for bitstream-corrupted video restoration. Corrupted frame sequences and masks are first fused with SAM2 structural embeddings and DINOv3 semantic tokens. A semantic-memory retrieval branch and an router-guided MoE-LoRA adaptation branch are jointly used for stage-1 content recovery. An optional stage-2 refinement module further suppresses residual artifacts and improves boundary consistency to produce the final recovered frame sequence.
Figure 6: The Vroom team: Enhanced B2SCVR: SAM2-Prior Guided Bitstream-Corrupted Video Restoration with LoRA and Boundary Refinement.
Figure 7: Overview of weichow team pipeline. The B2SCVR model processes frames at 432 × 240 (its native training resolution). Restored corrupted regions are upscaled to 1280 × 720 and composited with original uncorrupted pixels via binary masks. Non-corrupted frames bypass the model entirely for zero quality loss.
Figure 8: The holding team: Beyond Missing Holes: Taming Feature Leakage in Mask-Guided Bitstream-Corrupted Video Recovery. Built on a bidirectional propagation and temporal-transformer restoration backbone, our framework decomposes the input into valid and corrupted streams, suppresses unreliable corrupted-but-visible residuals with M1, retrieves cross-frame evidence through the target-centric cross-frame attention module (CA), and refines boundary seams with M2 after coarse restoration.
Bitstream-corrupted image restoration (BCIR) aims to reconstruct faithful images from corrupted compressed bitstreams. However, corruption occurs in bitstream-domain may lead to unpredictable error propagation and distortion amplification in pixel-domain, and even result in complete undecodability by standard decoders, posing significant challenge to existing image restoration paradigms. To address this challenge, we propose Bitstream Native Restorer (BiNar), a generative restoration framework that leverages fine-grained bitstream-native semantics to guide bitstream-corrupted image restoration. In BiNar, Bitstream Semantic Modeling (BSeM) extracts robust informative semantics and models sequential bytes directly from corrupted bitstream, enabling to generate fine-grained bitstream descriptions without relying on pixel decoding. Based on the descriptions,the Corruption-adaptive Multimodal Semantic Aggregation (CMSA) further extracts adaptive visual modal cues in decoded corrupted images and aggregates to bitstream descriptions. The multimodal aggregated fine-grained descriptions can serve as coherent restoration guidance. Finally, Blind Corruption Localization (BCoL) exploits the segmentation prior of visual foundation models to identify and locate error regions that requires recovery in corrupted images. The located error regions and the fine-grained descriptions jointly serve as visual prompt and language prompt of diffusion models image restoration, respectively. Extensive experiments under diverse bitstream corruption conditions show that BiNar outperforms all State-of-the-arts image restoration models regardless of perceptual quality or objective quality. Ablation studies demonstrate the effectiveness of fine-grained bitstream-native semantics in BCIR task.
This paper presents an overview of the NTIRE 2026 Challenge on Short-form UGC Video Restoration in the Wild with Generative Models. This challenge utilizes a new short-form UGC (S-UGC) video restoration benchmark, termed KwaiVIR, which is contributed by USTC and Kuaishou Technology. It contains both synthetically distorted videos and real-world short-form UGC videos in the wild. For this edition, the released data include 200 synthetic training videos, 48 wild training videos, 11 validation videos, and 20 testing videos. The primary goal of this challenge is to establish a strong and practical benchmark for restoring short-form UGC videos under complex real-world degradations, especially in the emerging paradigm of generative-model-based S-UGC video restoration. This challenge has two tracks: (i) the primary track is a subjective track, where the evaluation is based on a user study; (ii) the second track is an objective track. These two tracks enable a comprehensive assessment of restoration quality. In total, 95 teams have registered for this competition. And 12 teams submitted valid final solutions and fact sheets for the testing phase. The submitted methods achieved strong performance on the KwaiVIR benchmark, demonstrating encouraging progress in short-form UGC video restoration in the wild.
Bitstream-corrupted Harsh Visual Understanding (BcHVU) aims to understand harshly degraded videos originally decoded from a severely corrupted bitstream in real-world multimedia communication. The ill-posed nature of BcHVU poses a major challenge for existing vision models, as even subtle bitstream corruption can lead to irreversible pixel distortion and significant semantic loss. To address these challenges in BcHVU, we propose Bitstream Language Modeling as Robust Semantic Priors (BLMSP), a framework for learning and injecting bitstream-native semantic cues. Our proposed BLMSP framework learns to extract bitstream-native semantic cues by bitstream language modeling, and leverages them as priors by injecting into off-the-shelf vision models of BcHVU tasks. Specifically, we present a Video Bitstream Byte Model (VBBM) that integrates byte-level modeling and cross-codec semantic distillation, enabling it to interpret robust semantics from byte sequences in multiple corrupted bitstream formats. The learned bitstream semantics are leveraged as robust priors and fused into BcHVU model backbones for improving the quality of video restoration, captioning, and human pose estimation. To train BLMSP, we construct a large-scale multi-source Corrupted-bitstream Harsh-video Paired (CHP) dataset containing 607k corrupted bitstream segments and 287k paired harsh video clips. Extensive experimental results show that the learned bitstream priors improve video restoration, captioning, and human pose estimation by 2.51 dB in PSNR, 0.20 in CIDEr, and 0.18 in [email protected] on average, respectively. These results demonstrate that corrupted bitstream can serve as robust semantic priors in solving pixel distortion and semantic loss in BcHVU.