Alternating Gradient Descent and Mixture-of-Experts for Integrated Multimodal Perception

Hassan Akbari, Dan Kondratyuk, Yin Cui, Rachel Hornung, Huisheng Wang, Hartwig Adam

Keywords: Alternating Gradient Descent, Multimodal, Mixture of Experts, AGD, MoE, Deep Learning, Optimization

TL;DR: We combine Alternating Gradient Descent and Mixture-of-Experts to push the boundaries of efficient and scalable multimodal perception.

Abstract:

We present Integrated Multimodal Perception (IMP), a simple and scalable multimodal multi-task training and modeling approach. IMP integrates multimodal inputs including image, video, text, and audio into a single Transformer encoder with minimal modality-specific components. IMP makes use of a novel design that combines Alternating Gradient Descent (AGD) and Mixture-of-Experts (MoE) for efficient model & task scaling. We conduct extensive empirical studies and reveal the following key insights:
1) performing gradient descent updates by alternating on diverse modalities, loss functions, and tasks, with varying input resolutions, efficiently improves the model.
2) sparsification with MoE on a single modality-agnostic encoder substantially improves the performance, outperforming dense models that use modality-specific encoders or additional fusion layers and greatly mitigating the conflicts between modalities.
IMP achieves competitive performance on a wide range of downstream tasks including video classification, image classification, image-text, and video-text retrieval. Most notably, we train a sparse IMP-MoE-L focusing on video tasks that achieves new state-of-the-art in zero-shot video classification: 77.0% on Kinetics-400, 76.8% on Kinetics-600, and 68.3% on Kinetics-700, improving the previous state-of-the-art by +5%, +6.7%, and +5.8%, respectively, while using only 15% of their total training computational cost.

Submission Number: 13897