Camera-movement understanding model, supervised fine-tuned from Qwen/Qwen3-VL-8B-Instruct. Given a video, it outputs structured JSON describing every camera-movement segment — time span, basic-movement type / direction / speed, and special techniques.

Paper: Temporally Grounded Compositional Camera Motion Understanding via Geometric Knowledge Distillation
Project page: https://ddz16.github.io/cammotion.github.io
Code: https://github.com/ddz16/CamDistill

Usage

Recommended — the CamDistill repo

The CamDistill repo provides a one-line entry point that applies the official prompt and the exact video settings used for training and evaluation:

bash

python camera_movement_sft/infer_single.py \
    --model ddz16/CamSFT-8B \
    --video /path/to/video.mp4

With 🤗 Transformers

CamSFT is a standard Qwen3-VL model, so it can be loaded directly:

python

from transformers import Qwen3VLForConditionalGeneration, AutoProcessor

model = Qwen3VLForConditionalGeneration.from_pretrained(
    "ddz16/CamSFT-8B", dtype="bfloat16", device_map="auto"
)
processor = AutoProcessor.from_pretrained("ddz16/CamSFT-8B")

The exact system/user prompt and the video preprocessing (fps, max frames, resolution) are provided in the CamDistill repo; using them is required to reproduce the paper's results.