Skip to content

Sapere Aude

Gregor J. Rothfuss

Batch Normalization

we improve upon the best published result on ImageNet classification: reaching 4.9% top-5 validation error, exceeding the accuracy of human raters

Related

Uncategorized

Gregor J. Rothfuss
2015-02-22

ai
cs
google
science

Post navigation

The way of the dodo

Leave a comment

Δ

advertising (77) ai (238) algorithm (83) analysis (295) apple (94) archaeology (120) architecture (218) art (291) astronomy (145) augmentedreality (89) beer (80) biology (271) biotech (219) blogs (208) books (280) brain (155) business (238) china (189) climate (182) collaborative (99) comics (103) covid (127) crime (199) cs (80) culture (472) design (192) disaster (96) dna (164) economics (420) education (116) energy (142) environment (76) events (154) evolution (135) exploration (121) failedstate (153) fashion (80) finance (118) food (372) friends (162) funny (1058) games (242) google (346) googleearth (80) googlemaps (220) graphics (81) hardware (104) history (557) images (1648) innovation (161) internet (420) japan (79) languages (153) locative (97) mapping (260) math (81) me (667) media (241) medicine (348) microsoft (148) military (96) mobile (118) movies (150) music (212) nyc (822) opensource (177) performance (76) photography (134) physics (225) policy (223) politics (335) privacy (80) productivity (131) programming (210) psychology (165) realestate (96) religion (156) robotics (165) safety_fetish (82) science (1393) scifi (122) search (155) security (273) socialnetworks (125) society (225) software (183) space (221) standards (119) startup (92) stupid (511) switzerland (92) tourism (151) transit (87) transportation (161) ui (83) urbanism (138) us (363) video (1375) visualization (258) web (84)

Youtube
LinkedIn
Google Scholar

Comment
Reblog
Subscribe Subscribed
- Gregor J. Rothfuss
- Already have a WordPress.com account? Log in now.