Practical parallelism
System enables large speedups — as much as 88-fold — on common parallel-computing algorithms.
Computer Science and Artificial Intelligence Laboratory (CSAIL)
Peering into neural networks
New technique helps elucidate the inner workings of neural networks trained on visual data.
Computer system predicts products of chemical reactions
Machine learning approach could aid the design of industrial processes for drug manufacturing.
Drones that drive
CSAIL team’s system of quadcopters that fly and drive suggest another approach to developing flying cars.
Origami anything
New algorithm generates practical paper-folding patterns to produce any 3-D structure.
New technique makes brain scans better
Boosting quality of patient MRIs could enable large-scale studies of stroke outcome.
Shrinking data for surgical training
Technique that reduces video files to one-tenth their initial size enables speedy analysis of laparoscopic procedures.
Giving robots a sense of touch
GelSight technology lets robots gauge objects’ hardness and manipulate small tools.
Danielle Olson: Building empathy through computer science and art
CSAIL PhD student creates immersive media to help users understand each other’s backgrounds and feelings.
Armando Solar-Lezama: Academic success despite an inauspicious start
Initially turned down by nearly every college he applied to, the computer scientist cleared his own path to MIT.
Zika virus likely circulated in Americas long before its detection
Analysis of largest collection of Zika genomes to date reveals trajectory and evolution of the virus.
Using Bitcoin to prevent identity theft
System piggybacks on the digital currency’s security protocols to thwart hijacked servers.
New funding enables work on Internet policy and cybersecurity for key infrastructure
Work made possible by grants from the Internet Policy Research Initiative.
Cinematography on the fly
System directs camera-equipped drones to maintain framing of an aerial shot.
Teaching robots to teach other robots
CSAIL approach allows robots to learn a wider range of tasks using some basic knowledge and a single demo.