Text related annotation
Review related labels together against one full meeting transcript.
Human review of Hindi QA prelabels, source evidence, topics, and answer edits.
Read and record Egyptian Arabic questions for the SemEval 2027 multimodal dataset.
Read and record Levantine Arabic questions for the SemEval 2027 multimodal dataset.
Review Bangla open-ended questions and answers for the Bangladesh multimodal QA dataset.
Review Urdu open-ended questions and answers for the Pakistan multimodal QA dataset.
Review Turkish open-ended questions and answers for the Turkey multimodal QA dataset.
Human review of Assamese QA prelabels, source evidence, topics, and answer edits.
Human review of English QA prelabels, source evidence, topics, and answer edits.
In this task, you will help create high-quality Arabic conversation data. You will be shown a short multi-turn conversation where the ASSISTANT text is provided for context, and three USER turns are provided as text for you to speak.
Multimodal Natural Question, Voice Recording -- English Palestine
Human review of Bangla QA prelabels, source evidence, topics, and answer edits.
Read and record 1,000 Qatari Arabic sentences for the Mezza ASR dataset.
Read and record Bangla questions for the Bangladesh FinalSet dataset.
Read and record Chinese questions for the China FinalSet dataset.
Read and record Amharic questions for the Ethiopia FinalSet dataset.
Read and record Oromo questions for the Ethiopia FinalSet dataset.