Pith. sign in

REVIEW 21 cited by

Tarsier2: Advancing Large Vision-Language Models from Detailed Video Description to Comprehensive Video Understanding

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 2501.07888 v3 pith:WDZOQVVN submitted 2025-01-14 cs.CV cs.AI

Tarsier2: Advancing Large Vision-Language Models from Detailed Video Description to Comprehensive Video Understanding

classification cs.CV cs.AI
keywords videotarsier2-7bdetailedgpt-4otarsier2vision-languagedatadescription
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

We introduce Tarsier2, a state-of-the-art large vision-language model (LVLM) designed for generating detailed and accurate video descriptions, while also exhibiting superior general video understanding capabilities. Tarsier2 achieves significant advancements through three key upgrades: (1) Scaling pre-training data from 11M to 40M video-text pairs, enriching both volume and diversity; (2) Performing fine-grained temporal alignment during supervised fine-tuning; (3) Using model-based sampling to automatically construct preference data and applying DPO training for optimization. Extensive experiments show that Tarsier2-7B consistently outperforms leading proprietary models, including GPT-4o and Gemini 1.5 Pro, in detailed video description tasks. On the DREAM-1K benchmark, Tarsier2-7B improves F1 by 2.8% over GPT-4o and 5.8% over Gemini-1.5-Pro. In human side-by-side evaluations, Tarsier2-7B shows a +8.6% performance advantage over GPT-4o and +24.9% over Gemini-1.5-Pro. Tarsier2-7B also sets new state-of-the-art results across 15 public benchmarks, spanning tasks such as video question-answering, video grounding, hallucination test, and embodied question-answering, demonstrating its versatility as a robust generalist vision-language model.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 21 Pith papers

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

  1. MoHallBench: A Benchmark for Motion Hallucination in Video Large Language Models

    cs.CV 2026-07 unverdicted novelty 7.0

    MoHallBench is a new benchmark evaluating motion hallucination in VideoLLMs from co-occurrence priors, sequential inference, and similarity confusion, revealing decoupling from action recognition performance.

  2. CapRL++: Unified Reinforcement Learning with Verifiable Rewards for Dense Image and Video Captioning

    cs.CV 2026-06 unverdicted novelty 7.0

    CapRL++ applies reinforcement learning with verifiable rewards to dense image and video captioning by scoring captions via the accuracy of a vision-free LLM answering MCQs from the caption alone.

  3. TOC-Bench: A Temporal Object Consistency Benchmark for Video Large Language Models

    cs.CV 2026-05 conditional novelty 7.0

    TOC-Bench is a new diagnostic benchmark that reveals major weaknesses in temporal object consistency for Video-LLMs, including event counting, ordering, identity reasoning, and hallucination avoidance.

  4. TOC-Bench: A Temporal Object Consistency Benchmark for Video Large Language Models

    cs.CV 2026-05 unverdicted novelty 7.0

    TOC-Bench is an object-track-grounded benchmark that filters for temporally dependent questions and shows Video-LLMs have major weaknesses in event counting, ordering, identity reasoning, and hallucination detection.

  5. VideoChat3: Fully Open Video MLLM for Efficient and Generalist Video Understanding

    cs.CV 2026-07 conditional novelty 6.0

    An open 4B video MLLM with inflated-3D ViT tokenization and adaptive streaming perception outperforms comparable open models on general, long-video, and streaming benchmarks while using fewer visual tokens.

  6. LookME: Lookup-Based Multimodal Embeddings for Layer Injection in Vision-Language Models

    cs.CV 2026-07 conditional novelty 6.0

    LookME extends per-layer embedding lookup from text to continuous image and cross-modal embeddings and reports improved VLM scores on most benchmarks.

  7. MotionAtlas: Detailed Region Captioning for Motion-Centric Videos

    cs.CV 2026-06 unverdicted novelty 6.0

    MotionAtlas supplies a 2,073-question benchmark, a self-bootstrap pipeline yielding 159k captions, and fine-tuned Video-MLLMs that deliver 5.2-point gains over Qwen3-VL-4B on motion tasks.

  8. CineCap: Structured Reasoning with Spatio-Temporal Anchors for Cinematographic Video Captioning

    cs.AI 2026-06 unverdicted novelty 6.0

    CineCap combines structured reasoning and RL rewards to outperform baselines on cinematographic video captioning using a new 472-pair benchmark.

  9. Task-Focused Memorization for Multimodal Agents

    cs.CV 2026-05 unverdicted novelty 6.0

    TaskMem uses RL in two phases to learn a task-focused memorization policy for multimodal agents, yielding 5.3-7.0% VQA accuracy gains on reformulated streaming benchmarks from VideoMME, EgoLife, and EgoTempo.

  10. Syn4D: A Multiview Synthetic 4D Dataset

    cs.CV 2026-05 conditional novelty 6.0

    Syn4D supplies multiview synthetic dynamic scenes with dense geometric, tracking and pose ground truth that lets any pixel be unprojected to any time and camera.

  11. Kwai Keye-VL 1.5 Technical Report

    cs.CV 2025-09 conditional novelty 6.0

    Keye-VL-1.5 combines similarity-based Slow-Fast video token allocation with progressive context extension and iterative RL, reporting leading video-understanding results among 8B-scale multimodal models.

  12. V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning

    cs.AI 2025-06 unverdicted novelty 6.0

    V-JEPA 2 pre-trained on massive unlabeled video achieves strong results on motion understanding and action anticipation, SOTA video QA at 8B scale, and enables zero-shot robotic planning on Franka arms using only 62 h...

  13. VCap: Hypergeometric Rewards for Weak-to-Strong Visual Captioning

    cs.CV 2026-05 unverdicted novelty 5.0

    VCap pairs reference captions as witnesses with visual signals as adjudicators to deliver hypergeometric-precision rewards for RL in visual captioning, enabling an 8B model to outperform SOTA on benchmarks and improve...

  14. GazeBehavior Annotation Toolkit (GBAT): AI-powered toolkit for automatic annotation of egocentric eye-tracking and video data of child-caregiver interaction

    cs.CV 2026-05 unverdicted novelty 5.0

    GBAT is an AI toolkit that automates synchronization, gaze annotation, and action categorization for egocentric eye-tracking videos of early childhood interactions.

  15. Syn4D: A Multiview Synthetic 4D Dataset

    cs.CV 2026-05 unverdicted novelty 5.0

    Syn4D is a new multiview synthetic 4D dataset supplying dense ground-truth annotations for dynamic scene reconstruction, tracking, and human pose estimation.

  16. JoVA: Unified Multimodal Learning for Joint Video-Audio Generation and Editing

    cs.CV 2025-12 conditional novelty 5.0

    A dual-branch diffusion transformer with joint video-audio self-attention and a keypoint-based mouth-area loss reports top lip-sync and speech metrics on two benchmarks.

  17. NeMo: Needle in a Montage for Video-Language Understanding

    cs.CV 2025-09 conditional novelty 5.0

    NeMoBench, an automatically generated benchmark with 31,378 QA pairs, shows that video LLMs struggle with temporal grounding of relevant clips hidden in long montages.

  18. Temporal and Cross-Modal Alignment for Enhanced Audiovisual Video Captioning

    cs.CV 2026-07 unverdicted novelty 4.0

    TCA-Captioner introduces an Observer-Checker-Corrector refinement loop and TCA-Bench to address modality detachment and temporal incoherence in audiovisual video captioning.

  19. Watch, Remember, Reason: Human-View Video Understanding with MLLMs

    cs.CV 2026-06 unverdicted novelty 4.0

    This is a survey that frames video MLLM research via a human-view formulation of perceptual representations, memory states, reasoning traces, and predictions, then reviews methods, datasets, benchmarks, and open problems.

  20. LLaVA-OneVision-2: Towards Next-Generation Perceptual Intelligence

    cs.CV 2026-05 unverdicted novelty 4.0

    LLaVA-OV-2 uses codec-stream tokenization and a shared 3D RoPE to improve video, spatial, and tracking performance over Qwen3-VL-8B, while introducing the JumpScore benchmark for fine-grained motion localization.

  21. Seedance 1.0: Exploring the Boundaries of Video Generation Models

    cs.CV 2025-06 unverdicted novelty 4.0

    Seedance 1.0 generates 5-second 1080p videos in about 41 seconds with claimed superior motion quality, prompt adherence, and multi-shot consistency compared to prior models.