Hydra Attention: Efficient Attention with Many Heads
This addresses the computational bottleneck for applying transformers to large images in vision tasks, offering an incremental improvement in efficiency.
The paper tackles the computational inefficiency of self-attention in Vision Transformers for large images by introducing Hydra Attention, which uses as many heads as features to achieve linear scaling in tokens and features, resulting in significantly faster performance (by a factor of the token count) while maintaining or improving accuracy on ImageNet.
While transformers have begun to dominate many tasks in vision, applying them to large images is still computationally difficult. A large reason for this is that self-attention scales quadratically with the number of tokens, which in turn, scales quadratically with the image size. On larger images (e.g., 1080p), over 60% of the total computation in the network is spent solely on creating and applying attention matrices. We take a step toward solving this issue by introducing Hydra Attention, an extremely efficient attention operation for Vision Transformers (ViTs). Paradoxically, this efficiency comes from taking multi-head attention to its extreme: by using as many attention heads as there are features, Hydra Attention is computationally linear in both tokens and features with no hidden constants, making it significantly faster than standard self-attention in an off-the-shelf ViT-B/16 by a factor of the token count. Moreover, Hydra Attention retains high accuracy on ImageNet and, in some cases, actually improves it.