4 ms·
Do they need a pre-trained modality-specific model for each modality to train this?
by macleginn 5y ago
Do they need a pre-trained modality-specific model for each modality to train this?
- WithinReason 5y ago"Our work does not perform multimodal training but aims to unifiy (sic) the learning objective for self-supervised learning in different modalities."
- lopuhin 5y agoNo, they don't need a pre-trained model, they start from scratch. Also it's only the approach which is common between modalities, they train them separately.