Speech translation method, device and equipment and readable storage medium

The invention discloses a speech translation method, device and equipment and a readable storage medium. According to the scheme, after the to-be-translated source language speech is obtained, the source language speech is processed to obtain the source language text representation, the source language text representation is decoded to obtain the target language text, and in the scheme, the generated source language text feature is consistent with the sequence length of the real source language text, so that the target language text is obtained. Compared with the target language text obtained based on the acoustic features of the source language speech, the target language text obtained based on the source language text representation is higher in accuracy, so that the speech translation effect can be effectively improved by the scheme. 本申请公开了一种语音翻译方法、装置、设备及可读存储介质。本方案中，在获取待翻译的源语言语音之后，对源语言语音进行处理，得到源语言类文本表征，并对源语言类文本表征进行解码得到目标语言文本，该方案中，生成的源语言类文本特征与真实的源语言文本的序列长度一致，基于源语言类文本表征得到的目标语言文本相对于基于源语言语音的声学特征得到的目标语言文本准确度更高，