Everything At Once -- Multi-modal Fusion Transformer For Video Retrieval

Abstract

Multi-modal learning from video data has seen increased attention recently as it allows to train semantically meaningful embeddings without human annotation enabling tasks like zero-shot retrieval and classification. In this work, we present a multi-modal, modality agnostic fusion transformer approach that learns to exchange information between multiple modalities, such as video, audio, and text, and integrate them into a joined multi-modal representation to obtain an embedding that aggregates multi-modal temporal information. We propose to train the system with a combinatorial loss on everything at once, single modalities as well as pairs of modalities, explicitly leaving out any add-ons such as position or modality encoding. At test time, the resulting model can process and fuse any number of input modalities. Moreover, the implicit properties of the transformer allow to process inputs of different lengths. To evaluate the proposed approach, we train the model on the large scale HowT

Stats

• citations126
• S2 citations—
• github stars0
• HF likes0
• heat score15.78
• arxiv keyshvetsova2021everything

Related papers

• Multi-modal Transformer For Video Retrieval (2020)19.47
• MDMMT-2: Multidomain Multimodal Transformer For Video Retrieval, One More Step Towards Generalization (2022)0.00
• Joint Fusion And Encoding: Advancing Multimodal Retrieval From The Ground Up (2025)0.00
• Towards Fast Adaptation Of Pretrained Contrastive Models For Multi-channel Video-language Retrieval (2022)7.50
• Hit: Hierarchical Transformer With Momentum Contrast For Video-text Retrieval (2021)15.98
• Thinking Fast And Slow: Efficient Text-to-visual Retrieval With Transformers (2021)15.16
• Video And Audio Are Images: A Cross-modal Mixer For Original Data On Video-audio Retrieval (2023)7.16
• Unified Interactive Multimodal Moment Retrieval Via Cascaded Embedding-reranking And Temporal-aware Score Fusion (2025)0.00