OmniVec2 -- A Novel Transformer based Network for Large Scale Multimodal and Multitask Learning
This work addresses the problem of integrating diverse data types for AI applications, though it appears incremental as it builds on existing multimodal transformer approaches.
The authors tackled the challenge of large-scale multimodal and multitask learning by developing a transformer-based network that processes data from 12 modalities, achieving state-of-the-art performance across 25 datasets.
We present a novel multimodal multitask network and associated training algorithm. The method is capable of ingesting data from approximately 12 different modalities namely image, video, audio, text, depth, point cloud, time series, tabular, graph, X-ray, infrared, IMU, and hyperspectral. The proposed approach utilizes modality specialized tokenizers, a shared transformer architecture, and cross-attention mechanisms to project the data from different modalities into a unified embedding space. It addresses multimodal and multitask scenarios by incorporating modality-specific task heads for different tasks in respective modalities. We propose a novel pretraining strategy with iterative modality switching to initialize the network, and a training algorithm which trades off fully joint training over all modalities, with training on pairs of modalities at a time. We provide comprehensive evaluation across 25 datasets from 12 modalities and show state of the art performances, demonstrating the effectiveness of the proposed architecture, pretraining strategy and adapted multitask training.