MagicData

[Paper] [Project Page] [Github] 🧙 MagicData Dataset 1. Dataset Introduction The MagicData Dataset, introduced in MagicMotion, is a comprehensive trajectory controllable video generation dataset with rich semantic annotations. This dataset features high-quality videos with precise segmentation masks and bounding boxes

Papers using MagicData (4)

Leveraging Acoustic Contextual Representation by Audio-textual Cross-modal Learning for Conversational ASR2022Pronunciation-aware unique character encoding for RNN Transducer-based Mandarin speech recognition2022Speaker-Smoothed kNN Speaker Adaptation for End-to-End ASR2024Indonesian Automatic Speech Recognition with XLSR-532023