Research map / Video / Video generation
Test-Time Noise Guided Adaptation for Realistic Autoregressive Video Generation
Dimitrios Karageorgiou, Symeon Papadopoulos, Ioannis Kompatsiaris, Efstratios Gavves
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add test-time noise guided optimization (TANGO) | SelfForcing · VBench (15 s) | Total score: 0.827 -> 0.863 | Improved |
| add noise guided optimization (ablation) | SelfForcing · LV-Bench FVD | FVD: 576.8 -> 413.3 | Improved |
| add higher-order moment loss L_b (ablation) | TANGO · LV-Bench FVD | FVD: 506.8 -> 413.3 | Improved |
| add spectral flatness loss L_s (ablation) | TANGO · LV-Bench FVD | FVD: 487.9 -> 413.3 | Improved |
| add low-frequency moment loss L_l (ablation) | TANGO · LV-Bench FVD | FVD: 486.3 -> 413.3 | Improved |
| add first and second moment loss L_a (ablation) | TANGO · LV-Bench FVD | FVD: 441.6 -> 413.3 | Improved |
| add regularization L_reg (ablation) | TANGO · LV-Bench FVD | FVD: 664.2 -> 413.3 | Improved |
| increase LoRA rank from 2 to 8 (ablation) | TANGO · LV-Bench FVD | FVD: 596.2 -> 413.3 | Improved |
| increase LoRA rank from 8 to 16 (ablation) | TANGO · LV-Bench FVD | FVD: 413.3 -> 425.6 | No clear effect |
| increase number of noise vectors B from 1 to 10 (ablation) | TANGO · LV-Bench FVD | FVD: 473.7 -> 413.3 | Improved |
| increase B from 10 to 20 (ablation) | TANGO · LV-Bench FVD | FVD: 413.3 -> 399.4 | Improved |
| use lambda_reg 10 instead of 0.9 (ablation) | TANGO · LV-Bench FVD | FVD: 413.3 -> 475.6 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- RollingForcing — Rolling Forcing: Autoregressive Long Video Diffusion in Real Time (ICLR 2026) · better on 3 of 3 shared results
- LongLive — LongLive: Real-time Interactive Long Video Generation (ICLR 2026) · better on 3 of 3 shared results
- RewardForcing — Reward Forcing: Efficient Streaming Video Generation with Rewarded Distribution Matching Distillation (CVPR 2026) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Equilibrium Forcing: Adaptive Video Generation Without Noise Conditioning (NeurIPS 2026 · similarity 0.89)
- Self Forcing: Bridging the Train-Test Gap in Autoregressive Video Diffusion (NeurIPS 2025 · similarity 0.89)
- AR-Diffusion: Asynchronous Video Generation with Auto-Regressive Diffusion (CVPR 2025 · similarity 0.89)
- Identifying and Solving Conditional Image Leakage in Image-to-Video Diffusion Model (NeurIPS 2024 · similarity 0.88)
- Autoregressive Video Generation without Vector Quantization (ICLR 2025 · similarity 0.88)
- From Slow Bidirectional to Fast Autoregressive Video Diffusion Models (CVPR 2025 · similarity 0.88)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
