论文标题

Secemplip:将语音与预训练的视觉和语言模型相结合

SpeechCLIP: Integrating Speech with Pre-Trained Vision and Language Model

论文作者

Shih, Yi-Jen, Wang, Hsuan-Fu, Chang, Heng-Jui, Berry, Layne, Lee, Hung-yi, Harwath, David

论文摘要

数据驱动的语音处理模型通常在大量文本监督下表现良好,但是收集抄录的语音数据成本高昂。因此,我们提出了SpeechClip,这是一个新颖的框架,通过图像桥接语音和文本,以增强语音模型而无需转录。我们利用最先进的预训练的休伯特和剪辑,通过配对的图像和口语标题对齐,并以最小的微调对齐。 SpeechClip在图像语音检索上的先验优于先前的最新时间,并且在没有转录的直接监督的情况下执行零击的语音文本检索。此外,SecemClip可以直接从语音中检索语义上相关的关键字。

Data-driven speech processing models usually perform well with a large amount of text supervision, but collecting transcribed speech data is costly. Therefore, we propose SpeechCLIP, a novel framework bridging speech and text through images to enhance speech models without transcriptions. We leverage state-of-the-art pre-trained HuBERT and CLIP, aligning them via paired images and spoken captions with minimal fine-tuning. SpeechCLIP outperforms prior state-of-the-art on image-speech retrieval and performs zero-shot speech-text retrieval without direct supervision from transcriptions. Moreover, SpeechCLIP can directly retrieve semantically related keywords from speech.

扫码加入交流群

加入微信交流群

微信交流群二维码

扫码加入学术交流群,获取更多资源