Benchmark Almanac

MIT Computer Science

We track 2 benchmarks that MIT Computer Science made or helped make.

0 in use 0 with public code 0 with a leaderboard here

Benchmarks

Name Kind Released Best score Status
ADE20K Models label the pixels in a scene by object or region. Meta measures overlap between predicted and reference segments. Detection and captioning Aug 2016 Not in cards
Places205 Models classify photographs by scene type. Meta compares scene recognition on the 205-category edition of Places. General visual questions Not in cards