V²L: Leveraging Vision and Vision-language Models into Large-scale Product RetrievalShare on Twitter Facebook LinkedIn Previous Next