Speech recognition method and device

The invention provides a speech recognition method and device. The method comprises the following steps: dividing obtained sound data to be identified into a plurality of overlapped sound frames; carrying out fast Fourier transform operation on each sound frame to obtain corresponding frequency spectrum; carrying out traversal on the frequency spectrum of the plurality of overlapped sound frames, and dividing the traversed frequency spectrum of the current sound frame into a plurality of non-overlapped sub-bands; according to frequency spectrum energy of the plurality of sub-bands of the current sound frame, calculating energy mean square root of the current sound frame; and when determining that the energy mean square root of the current sound frame is larger than a preset threshold value, determining that the current sound frame comprises speech information. According to the scheme above, speech recognition speed and accuracy can be improved. 语音识别方法及装置，所述方法包括：将获取的待识别的声音数据划分为多个重叠的声音帧，并对每个声音帧进行快速傅立叶变换运算，得到对应的