PUBLICATIONS
CBB-FE, CamemBERT and BiT Feature Extraction for Multimodal Product Classification and Retrieva
Jul 2020
For SIGIReCom’20 multimodal product classification and retrieval challenge, we proposed a solution that consisting of (a) fine-tuning feature extractors from text and image modalities respectively, (b)fusing multimodal features to train a classifier for the first task on product classification, and (c) using a similarity search method to retrieve product images from their text titles. Utilizing latest
Copied!
Research Areas :
#Language Program
Careers :
Open Positions