So I’ve had a phone call this morning off my new(ex) employer basically saying my offer has been revoked. I’m devastated by this news, especially after working so hard to get to this point alongside working a minimum wage job. (1/2)

What media often portrays as a good scientist: -genius What it actually takes to be a good scientist: -curiosity -perseverance -teamwork -resilience -leadership

I propose we rename "epistemic uncertainty" to "model uncertainty", and "aleatoric uncertainty" to "per-measurement uncertainty". More generally, one can refer to uncertainty at any level, such as per-pixel, per-image, or per-patient uncertainty. No need for obscure jargon.

Our paper “Drop-DTW: Aligning Common Signal Between Sequences While Dropping Outliers” is accepted at #NeurIPS2021! Joint work with Nikita isma me Animesh Garg and Allan Jepson. Shout-out to the reviewers for holding a constructive dialog.

**MBT: New multimodal fusion architecture** Versatile self-attention based framework for any number/type of modality! The paper “Attention Bottlenecks for Multimodal Fusion” has been accepted to #NeurIPS2021. Link: arxiv.org/abs/2107.00135 w Google AI

Thinking about grad school next year? Interested in visualization, human/AI interaction, computational art? Consider Harvard! Martin Wattenberg and I are building up our lab and will be looking for students. Our DMs are open.

#atoMeow stippling demo code for #processing released @Opensource GitHub GitHub: github.com/mwkm/atoMeow Can't wait to see your #atomMeow stipples Happy 20th Anniversary Processing #dotdot #bluenoise #creativecoding #LagrangianImage #p5js #digitalart #generativeart

My lab is looking for two talented PhD students in topics related to egocentric vision: Topic 1: "Long Range Video Understanding" Topic 2: "Ambient Aware Wearable Vision for Energy Saving in Industrial Sites" Call: unict.it/sites/default/… Deadline 15/10/2021

Which Shortcut Cues Will DNNs Choose? A Study from the Parameter-Space Perspective Luca Scimeca, Seong Joon Oh, SanghyukChun, Michael Poli, Sangdoo Yun tl;dr: provide CNN entangles features, check which it uses for prediction. Color & ethnicity rules. arxiv.org/abs/2110.03095…

Any #xmonad users, you can sponsor the project here: github.com/sponsors/xmonad

We’re announcing #Ego4D, an ambitious long-term project we’ve embarked on w/13 universities in 9 countries to advance first-person perception. This work will catalyze research to build more useful AI assistants, robots & other future innovations. ai.facebook.com/blog/teaching-…

Read about Bristol's part in the #Ego4D project through our press release @BristolUniEng bristol.ac.uk/news/2021/octo… Check Interactive Webpage: ego4d-data.org Big shoutout to team: Michael Wray, Will Price, Jonathan Munro and Adriano Fragomeni

CNNs ask you to predefine the size of conv. kernels. But, isn't it better to let the network decide? Our new work w/ @rjbruin let CNNs learn full-res. kernels during training, leading to SOTA on several datasets!😀 Code & slides here! w/ Jakub Tomczak Erik Bekkers & @jan_gemert

image-scaling attacks are wild small dots added to the image on the left turns it into the image on the right when downscaled could make auditing ML systems very tricky if you only look at the original images...

Amazing move British Machine Vision Conference (BMVC) The BMVA for accepted papers, registration fee of £42 to cover server hosting &fixed costs for other attendees **registration is free** unwind after #CVPR2021 deadline by attending British Machine Vision Conference (BMVC) for free 22-25 Nov. Kudos to PCs Neill Campbell Karteek Alahari

Congratulations to Simon Peyton Jones who is joining Epic Games to work with Lennart Augustsson on the Verse language for game design and programming the metaverse! discourse.haskell.org/t/an-epic-futu…

Presenting SEAL: Self-supervised Embodied Active Learning! #NeurIPS2021 SEAL is a self-supervised framework to close the action-perception loop. It improves perception & action models by just moving in the physical world w/o any human supervision. devendrachaplot.github.io/projects/seal 1/N

Major Congrats to ***Dr Will Price*** Will Price who just passed his viva on "The Role of Time in Video Understanding" with no corrections @BristolUniEng Thanks to Prof Yoichi Sato (#UTokyo) and Laurence Aitchison for examining the thesis. 1/N

Lambda Calculus in 400 Bytes by Justine Tunney is a top-notch semester-long university computer science course in a single page justine.lol/lambda/