AI model speeds up high-resolution computer vision

Researchers from MIT and the MIT-IBM Watson AI Lab have developed an efficient computer vision model for autonomous vehicles and other high-resolution computer vision tasks. This model, called EfficientViT, performs real-time semantic segmentation with linear computational complexity, outperforming previous models by running up to nine times faster while maintaining accuracy. It simplifies the attention map construction and balances performance and efficiency by adding components for capturing local features and multiscale learning. The model’s hardware-friendly design makes it adaptable to various devices, including autonomous vehicles, virtual reality headsets, and edge computers. It holds the potential to improve real-time decision-making and efficiency