The challenge is divided into two tracks: an image track and an audio-video track. The competition provides high-quality data sets and model training resources, which can comprehensively and fairly test the participating teams when facing complex and changeable scenes, a wide range of crowd characteristics, and different generation techniques. At the same time, a million-scale prize pool is set up to encourage scholars, engineers, educators, students, and independent developers to actively participate in the challenge.
This is a technical competition, and it is also an exploration of the boundaries of information security in the intelligent era. We sincerely invite all parties to join us for the feast of intelligent competition and jointly protect the digital future of authenticity and trust.