Pith. sign in

REVIEW 7 cited by

LiFT: Leveraging Human Feedback for Text-to-Video Model Alignment

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2412.04814 v3 pith:UBU3DY5P submitted 2024-12-06 cs.CV

classification cs.CV
keywords humanmodelalignmentfeedbackmodelsrewardvideosaligning
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Recent advances in text-to-video (T2V) generative models have shown impressive capabilities. However, these models are still inadequate in aligning synthesized videos with human preferences (e.g., accurately reflecting text descriptions), which is particularly difficult to address, as human preferences are subjective and challenging to formalize as objective functions. Existing studies train video quality assessment models that rely on human-annotated ratings for video evaluation but overlook the reasoning behind evaluations, limiting their ability to capture nuanced human criteria. Moreover, aligning T2V model using video-based human feedback remains unexplored. Therefore, this paper proposes LiFT, the first method designed to leverage human feedback for T2V model alignment. Specifically, we first construct a Human Rating Annotation dataset, LiFT-HRA, consisting of approximately 10k human annotations, each including a score and its corresponding rationale. Based on this, we train a reward model LiFT-Critic to learn reward function effectively, which serves as a proxy for human judgment, measuring the alignment between given videos and human expectations. Lastly, we leverage the learned reward function to align the T2V model by maximizing the reward-weighted likelihood. As a case study, we apply our pipeline to CogVideoX-2B, showing that the fine-tuned model outperforms the CogVideoX-5B across all 16 metrics, highlighting the potential of human feedback in improving the alignment and quality of synthesized videos.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 7 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Temporal Concentration from Rollout Errors: Implicit Preference Optimization for Text-to-Video Diffusion

    cs.CV 2026-07 conditional novelty 6.0 of 10

    Implicit DPO pairs from real-vs-reconstruction rollouts plus concentration on high latent-error temporal windows improve video authenticity and coherence without annotations or reward models.

  2. Reward Lightning: Fast Video Generation via Homologous Preference Distillation

    cs.CV 2026-07 conditional novelty 6.0 of 10

    Homologous preference distillation evaluates adversarial distillation and latent reward alignment on identical latent features, yielding 1–4-step video generators that improve VBench by 2.1% while leading text, motion...

  3. FantasyTalking2: Timestep-Layer Adaptive Preference Optimization for Audio-Driven Portrait Animation

    cs.CV 2025-08 unverdicted novelty 6.0 of 10

    A three-part system, Talking-Critic, Talking-NSQ, and TLPO, aligns diffusion portrait animation models to human preferences and improves lip-sync, motion naturalness, and visual quality.

  4. VQ-Insight: Teaching VLMs for AI-Generated Video Quality Understanding via Progressive Visual Reinforcement Learning

    cs.CV 2025-06 conditional novelty 6.0 of 10

    VQ-Insight uses progressive reinforcement learning with temporal shuffle and task rewards to teach a vision-language model to score and compare AI-generated videos, with gains on multiple video quality benchmarks.

  5. VCapsBench: A Large-scale Fine-grained Benchmark for Video Caption Quality Evaluation

    cs.CV 2025-05 conditional novelty 6.0 of 10

    VCapsBench is a video caption quality benchmark with 109,796 QA pairs across 21 fine-grained dimensions on 5,677 videos, evaluating caption accuracy, inconsistency, and coverage.

  6. GigaVideo-1: Advancing Video Generation via Automatic Feedback with 4 GPU-Hours Fine-Tuning

    cs.CV 2025-06 conditional novelty 5.0 of 10

    GigaVideo-1 fine-tunes Wan2.1 on synthetic weakness-targeted prompts with VLM reward reweighting and reports ~4% average VBench-2.0 gains per dimension at 4 GPU-hours each, though joint training gains less.

  7. Reward Modeling for Reinforcement Learning-Based LLM Reasoning: Design, Challenges, and Evaluation

    cs.LG 2026-02 conditional novelty 4.0 of 10

    A taxonomy-driven survey arguing that reward design is the central mechanism shaping reliable LLM reasoning, with maps of reward paradigms, reward-hacking failure modes, and benchmark pitfalls.

Pith tools