# Eyeline-Labs/Go-with-the-Flow

The official implementation of CVPR'25 Oral paper "Go-with-the-Flow: Motion-Controllable Video Diffusion Models Using Real-Time Warped Noise"

Repository: https://github.com/Eyeline-Labs/Go-with-the-Flow
Canonical: https://ross.abutalabs.com/products/go-with-the-flow
Homepage: https://eyeline-labs.github.io/Go-with-the-Flow/
Language: Python
License: NOASSERTION
License Family: other
Last push: 2025-10-13T17:04:31+00:00

## Health v2 (maintenance only)
Score: 41/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 46, release rhythm 35, longevity 42
- inputs: {"age_days": 601, "days_push": 324, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1093, forks 50 (observed 2026-08-28T04:03:33.742192+00:00)

## What it is
Official implementation of the CVPR 2025 Oral paper 'Go-with-the-Flow', which controls motion in video diffusion models by replacing i.i.d. Gaussian noise with warped noise during fine-tuning and inference. It includes a local GUI for cut-and-drag motion specification and scripts to run image-to-video and text-to-video diffusion with controllable camera and object motion.

## Use cases
- generate videos with controlled camera motion
- make a still image move with cut-and-drag animation
- transfer motion patterns from one video to another
- control object motion in video diffusion model output
- run text-to-video generation with a motion prior
- animate a segmented object in an image

## When to choose
- you need fine-grained control over motion in video diffusion outputs
- you want to animate a still image by dragging objects or specifying camera paths
- you want to transfer motion from a reference video to new content
- you are reproducing or building on the CVPR 2025 warped-noise paper

## When to avoid
- you need a polished production video-generation product rather than research code
- you have no GPU available for diffusion inference
- you need a non-Python or fully headless pipeline without the GUI
- you require a permissive license - the license is custom (NOASSERTION)

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, video-processing, image-processing, llm-inference
- domain: deep-learning, computer-vision, image-processing, artificial-intelligence
- platform: python
- tags: video-diffusion, motion-control, warped-noise, generative-ai, text-to-video, image-to-video, cvpr-2025, research-code, video, gpu, linux

## Member repositories
- Eyeline-Labs/Go-with-the-Flow (main) score 41

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:33.742192+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:47:39.348169+00:00, confidence not recorded.
  - readme: https://github.com/Eyeline-Labs/Go-with-the-Flow (fetched 2026-08-28T04:03:33.742192+00:00, sha d372195bd6a7)
  - homepage: https://eyeline-labs.github.io/Go-with-the-Flow/ (fetched 2026-08-29T12:50:39.758940+00:00, sha 89c96ac3f787)
- Data as of 2026-08-30T08:39:29.467469+00:00.