PulseAugur
EN
LIVE 08:07:21

Goku dataset and benchmark advance instruction-based video editing · 4 sources tracked

By PulseAugur Editorial · [4 sources] ·

Researchers have introduced Goku, a large-scale dataset and benchmark for instruction-based video editing, designed to overcome the limitations of existing datasets that focus on single-task appearance edits. Goku comprises 2 million high-quality video editing pairs, enabling multi-task and structural manipulations like precise subject movement control. The accompanying Goku-Edit model utilizes a multimodal large language model for instruction comprehension and a dual-branch design for structural and appearance editing. A benchmark, Goku-Bench, with 1,000 human-verified cases and 7 new metrics, was also released, showing Goku-Edit achieving up to an 8% improvement in instruction following over other open-source models. AI

IMPACT Advances capabilities in instruction-based video editing, potentially enabling more complex and creative user-driven video manipulations.

RANK_REASON The cluster describes a new dataset, benchmark, and model for video editing, published as a research paper.

Read on Hugging Face Daily Papers →

AI-generated summary · Google Gemini · from 4 sources. How we write summaries →

Goku dataset and benchmark advance instruction-based video editing · 4 sources tracked

How we ranked this

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Research
The cluster describes a new dataset, benchmark, and model for video editing, published as a research paper.
Source corroboration
4 independent sources
Strong cross-source corroboration — multiple independent publishers covered this within the clustering window.
Topics
paper, product
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
57 days old
Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

Full methodology in our editorial standards.

COVERAGE [4]

  1. Hugging Face Daily Papers TIER_1 English(EN) ·

    Goku: A Million-Scale Universal Dataset and Benchmark for Instruction-Based Video Editing

    A large-scale video editing dataset and model are introduced that support multi-task and structural manipulations through advanced data synthesis and network architectures.

  2. Hugging Face Daily Papers TIER_1 English(EN) ·

    Goku: A Million-Scale Universal Dataset and Benchmark for Instruction-Based Video Editing

    Existing instruction-based video editing datasets commonly focus on single-task appearance editing, failing to meet the complex creative demands of real-world scenarios. To bridge this gap, we present Goku, a large-scale dataset featuring 2 million high-quality, instruction-align…

  3. arXiv cs.CV TIER_1 English(EN) · Sen Liang, Cong Wang, Zhentao Yu, Fengbin Guan, Zhengguang Zhou, Teng Hu, Youliang Zhang, Yuan Zhou, Xin Li, Qinglin Lu, Zhibo Chen ·

    Goku: A Million-Scale Universal Dataset and Benchmark for Instruction-Based Video Editing

    arXiv:2606.30599v1 Announce Type: new Abstract: Existing instruction-based video editing datasets commonly focus on single-task appearance editing, failing to meet the complex creative demands of real-world scenarios. To bridge this gap, we present Goku, a large-scale dataset fea…

  4. arXiv cs.CV TIER_1 English(EN) · Zhibo Chen ·

    Goku: A Million-Scale Universal Dataset and Benchmark for Instruction-Based Video Editing

    Existing instruction-based video editing datasets commonly focus on single-task appearance editing, failing to meet the complex creative demands of real-world scenarios. To bridge this gap, we present Goku, a large-scale dataset featuring 2 million high-quality, instruction-align…

RELATED ENTITIES

RELATED TOPICS