EVFHQ (Extended VFHQ) is an expansion of the VFHQ dataset, which offers a significantly larger and more diverse collection of high-quality video face data.
EVFHQ consists of the following Dockerized components:
- EVFHQ-Database: Manages a PostgreSQL database that stores video metadata and processing results.
- EVFHQ-Fetcher: Responsible for gathering video metadata from YouTube.
- EVFHQ-Assigner: Distributes video download tasks across multiple servers to optimize load and efficiency.
- EVFHQ-Downloader: Handles the downloading of video files from YouTube.
- EVFHQ-Processor (Coming Soon): Specializes in extracting facial clips from the downloaded videos.