To gain some perspective, try https://farm5.static.flickr.com/4096/4921383047_6e33b35d4e_b.jpg … on http://cloudcv.org/vqa/ and ask almost anything. #dlearn
-
-
Replying to @policecar
@policecar ok these were meanpic.twitter.com/cTJF13qoFB
2 replies 0 retweets 0 likes -
Replying to @allgebrah
@policecar interesting to play with its sense of numberspic.twitter.com/1xULObOKTu
1 reply 0 retweets 0 likes -
Replying to @allgebrah
@policecar I can't explain why it recognizes so few people, especially with those mirrorspic.twitter.com/MSQr3uA5vn
2 replies 0 retweets 0 likes -
Replying to @allgebrah
@allgebrah would be neat if it presented evidence for its decisions, wouldn’t it?1 reply 0 retweets 0 likes -
Replying to @policecar
@policecar as in "these blobs I consider X"? or something deeper?1 reply 0 retweets 0 likes -
Replying to @allgebrah
@allgebrah by tendency hard w/ neural networks but the convolutional part of the model has features that could be visualized1 reply 0 retweets 0 likes -
Replying to @policecar
@policecar I pass, I would need more than my superficial knowledge of convnets here :/2 replies 0 retweets 0 likes -
Replying to @allgebrah
@allgebrah i was thinking of visualizations as proposed in http://www.iro.umontreal.ca/~lisa/pointeurs/invariances_techreport.pdf … which would allow one to generate images that the ..1 reply 0 retweets 1 like
@policecar Will read, thanks!
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.