Sjasilva / VSFA

Quality Assessment of In-the-Wild Videos, accepted by ACM MM 2019

Quality Assessment of In-the-Wild Videos

Description

VSFA code for the following papers:

• Dingquan Li, Tingting Jiang, and Ming Jiang. Quality Assessment of In-the-Wild Videos. In Proceedings of the 27th ACM International Conference on Multimedia (MM ’19), October 21-25, 2019, Nice, France. [arxiv version]

Requirement

pip install -r requirements.txt -i https://pypi.tuna.tsinghua.edu.cn/simple

• PyTorch 1.1.0
• TensorboardX 1.2, TensorFlow-TensorBoard
• pytorch/ignite

Intra-Database Experiments (Training and Evaluating)

Feature extraction

CUDA_VISIBLE_DEVICES=0 python CNNfeatures.py --database=KoNViD-1k --frame_batch_size=64

You need to specify the database and change the corresponding videos_dir.

Quality prediction

CUDA_VISIBLE_DEVICES=0 python VSFA.py --database=KoNViD-1k --exp_id=0

You need to specify the database and exp_id.

Test Demo

python test_demo.py --video_path=test.mp4

Contact

Dingquan Li, dingquanli AT pku DOT edu DOT cn.