Page 6 - Mit Csail News Today

MIT PhD student Yilun Du describes methods to make generative art systems like DALL-E 2 more creative, while also exploring using stable diffusion models for robots.

Researchers from MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) created a new method for magnetically programming materials to make cubes that are very picky about who they connect with, enabling more scalable self-assembly.

Researchers from MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) created a system called MuscleRehab for visualizing muscle engagement and motion data during unsupervised physical rehabilitation.

SEED is a new soft-robotics system that can grasp tools and apply the appropriate amount of force for a given task, like squeegeeing up liquid or writing out a word with a pen. The work is a collaboration between MIT CSAIL and the Toyota Research Institute.

The MIT Lincoln Laboratory Supercomputing Center has released a dataset containing more than a million jobs run on its TX-GAIA supercomputer, with the goal of using this data to feed AI research into optimizing data center resources. The team has kicked off the MIT Datacenter Challenge to mobilize this research.

Amar Gupta of MIT's Computer Science and Artificial Intelligence Laboratory discusses the increase in virtual care driven by Covid-19 and the need for an integrated approach to enhanced care delivery.

Inspired by a fiddler crab eye, scientists developed an amphibious artificial vision system with a panoramic visual field. It is the first artificial vision system for both land and water.

A new framework called FabO uses on-screen visual content to integrate real-world fabrication with existing video games. FabO, developed by researchers at MIT CSAIL, aims to enliven the teaching of basic maker skills.

MIT CSAIL researchers developed a novel method to "see through" objects in a scene, and exploit the projection for 3D reconstruction and reasoning.

MIT scientists found a new hardware vulnerability in the Apple M1 chip. They created a novel PACMAN attack methodology to show that Pointer Authentication Code - the last line of defense against typical software vulnerabilities - can be defeated without leaving a trace.