【Product Features】High sampling rate, in-vehicle corpus, indoor quiet collection, multiple scenarios (vehicle control, music, general, map, casual chat scenarios) Can be applied to in-vehicle and other common speech recognition scenarios. 【Audio Parameters】 16k: 1 person 0.5 hours 44.1k: 148 people 74.9 hours 48k：2463 people,1354.8 hours