The Second DISPLACE challenge entails a first of kind task to perform speaker and language diarization as well as speech recognition on a demanding dataset, where the data contains multi-speaker social conversations in multilingual code-mixed speech. In such cases, we find that current speech processing systems  are not equipped to perform meaningfully. The first DISPLACE challenge in 2023 illustrated this abundantly -   https://arxiv.org/pdf/2311.12564.pdf

With this motivation, the Second DISPLACE challenge attempts to advance this further by benchmarking and improving Speaker Diarization (SD) in multilingual settings and Language Diarization (LD) in multi-speaker settings, using the same underlying dataset. Further, the track on speech recognition attempts to improve speech transcription in code-mixed multi-speaker speech.  Registrations are open for this challenge which will contain two tracks - a) Speaker diarization track , b) Language diarization track and c) ASR task. 

A baseline system and an open leaderboard is available to the participants. The DISPLACE challenge is split into two phases, where the first phase is linked to the Interspeech paper submission deadline, while the second phase aligns with the camera ready submission deadline. For more details, dates and to register, kindly visit the DISPLACE challenge website: https://displace2024.github.io

We look forward to your team challenging to 'displace' the state-of-the-art in speaker, language diarization systems and/or ASR systems. 

Thank you and Namaste,

The DISPLACE team