@dennybritz@deliprao@MelMitchell1 I agree with Denny. Self-supervision is about solving a pretext task which is defined by the data itself. For example intrinsic images could be learned by a time lapse videos without any extra label but just by utilizing the fact of changing illumination and constant reflectance
If you could represent a paper as a code diff, many papers could be compressed down to <50 lines :) The diff would also be more intuitive to read and eval standardized.
Some ideas are so different that this wouldn’t apply, but I think it would work well for the majority.
Just pushed the code of a chrome extension that turns every Instagram posts into 3d images using #3DPhotoInpainting. No GPU needed thanks to @GoogleColab but a bit of patience to set it up ;-)
Demo: @parrstudio's amazing work
Code: https://t.co/59yJUvRHxE
#AIUX#Interaction#ML
@askerlee@jbhuang0604@alexjc Maybe. Maybe Not? What if the model implicitly learned to recognize surface normals and perform transformations based on that?
Crowdsourced generative art gallery 👇🏽
This is art created and described by 66 anonymous individuals on Amazon Mechanical Turk. They used a “Create Your Own” tool from https://t.co/oC2Qm5Y1c2 to make these. I'll post one ~every week.
#generativeart#creativecoding#crowdsourced
@olivercameron Great Balanced Read! A minor hiccup, had to go through your Twitter threads to properly understand the paper in discussion. Summarizing those in the newsletter itself possible?
@emirates Baggage Allowance Extension Student Offer ends on April 30th! A lot of students are gonna receive their i-20 after that date (even though they applied a long time back) Do extend the date please.