Today's paper reading was a bit rushed and I posted about only 7 papers out of 79 in the cs.CV category on arXiv.

There weren't a lot of papers that interested me, but on the other hand, this could also be a product of me being tired and not finding a lot of things as interesting 😛

#AI #CV #NewPapers #DeepLearning #MachineLearning

"Input Perturbation Reduces Exposure Bias in Diffusion Models. (arXiv:2301.11706v1 [cs.LG])" — An exploration of the fact that the the long sampling chain in Denoising Diffusion Probabilistic Models (DDPM) leads to an error accumulation phenomenon, which is similar to the exposure bias problem in autoregressive text generation.

Paper: http://arxiv.org/abs/2301.11706

#AI #CV #NewPaper #DeepLearning #MachineLearning

<<Find this useful? Please boost so that others can benefit too 🙂>>
The inputs and the prediction t…

"Image Restoration with Mean-Reverting Stochastic Differential Equations. (arXiv:2301.11699v1 [cs.LG])" — A stochastic differential equation (SDE) approach for general-purpose image restoration which can restore images without relying on any task-specific prior knowledge.

Paper: http://arxiv.org/abs/2301.11699

#AI #CV #NewPaper #DeepLearning #MachineLearning

<<Find this useful? Please boost so that others can benefit too 🙂>>
An overview of our proposed con…

"Accelerating Guided Diffusion Sampling with Splitting Numerical Methods. (arXiv:2301.11558v1 [cs.CV])" — A solution to speeding up guided diffusion image generation based on operator splitting methods, motivated by the finding that classical high-order numerical methods are unsuitable for the conditional function.

Paper: http://arxiv.org/abs/2301.11558

#AI #CV #NewPaper #DeepLearning #MachineLearning

<<Find this useful? Please boost so that others can benefit too 🙂>>
Generated samples of a classifi…

"3DShape2VecSet: A 3D Shape Representation for Neural Fields and Generative Diffusion Models. (arXiv:2301.11445v1 [cs.CV])" — A novel shape representation for neural fields designed for generative diffusion models, which can encode 3D shapes given as surface models or point clouds, and represents them as neural fields.

Paper: http://arxiv.org/abs/2301.11445

#AI #CV #NewPaper #DeepLearning #MachineLearning

<<Find this useful? Please boost so that others can benefit too 🙂>>
Left: Shape autoencoding result…

"Improving Cross-modal Alignment for Text-Guided Image Inpainting. (arXiv:2301.11362v1 [cs.CV])" — A model for text-guided image inpainting by improving cross-modal alignment (CMA) using cross-modal alignment distillation and in-sample distribution distillation.

Paper: http://arxiv.org/abs/2301.11362

#AI #CV #NewPaper #DeepLearning #MachineLearning

<<Find this useful? Please boost so that others can benefit too 🙂>>
Different categories of vision-…

"Rethinking 1x1 Convolutions: Can we train CNNs with Frozen Random Filters?. (arXiv:2301.11360v1 [cs.CV])" — An exploration into whether Convolutional Neural Networks (CNN) learning the weights of vast numbers of convolutional operators is really necessary.

Paper: http://arxiv.org/abs/2301.11360

#AI #CV #NewPaper #DeepLearning #MachineLearning

<<Find this useful? Please boost so that others can benefit too 🙂>>
Validation accuracy of LCResNet…

"Multimodal Event Transformer for Image-guided Story Ending Generation. (arXiv:2301.11357v1 [cs.CV])" — A multimodal event transformer, an event-based reasoning framework for image-guided story ending generation which constructs visual and semantic event graphs from story plots and ending image, and leverages event-based reasoning to reason and mine implicit information in a single modality.

Paper: http://arxiv.org/abs/2301.11357

#AI #CV #NewPaper #DeepLearning #MachineLearning

<<Find this useful? Please boost so that others can benefit too 🙂>>
Given a multi-sentence story pl…

A total of 79 papers in the cs.CV category on arXiv.org today — 52 new, 27 updated.

Had a really long (or it felt long) weekend struggling with code. So would it be a relief to read papers? I don't know ... 😛

#AI #CV #NewPapers #DeepLearning #MachineLearning

Reminder: we are accepting submissions of extended abstracts for presentations at the Inaugural Conference of the Technology Ethics eXchange of the NorthEast (teχnē). This conference will take place at Harvard in May. Deadline is Feb 3.

Also announcing our keynote speakers! We will have talks from Helen Nissenbaum, Casey Fiesler @cfiesler, and Anita Allen!

More information:
https://philevents.org/event/show/105209

@philosophy @facct
#philosophy #TechEthics #AIethics #ResponsibleComputing #CFP #conference

By now you know that #ChatGPT will make up nonsense, presented with confidence.

Useful framework by AI and Data Policy lawyer:

Hello everyone! I just created my profile here, so this is my first post and of course, an #introduction . I'm still trying to understand how this social media works.
Anyway, my name is Matsya and i'm an artist living in Brazil. I work mostly traditionally with watercolors and inks and what drives me most is folktales, whimsical landscapes and fantasy characters.

#MastoArt #watercolor #watercolour #painting #art

Tahoma (Mt Rainier), Washington.

Taken near the Sunrise Lodge in the Mt Rainier Nat'l Park. Featuring a lenticular cloud forming over the crest, the Emmons Glacier and the White River.

#Photography #MtRainier #MtRainierNationalPark #WashingtonState

Liebe Grüße an Euch. 👋

#Nordlichter #Schnee #reflection #DigitalArt

Anyone know an artist that can do good 3d cutaways of space craft or military vehicles? I’m looking to pay to have one made for a new books series I have coming out. I’d like to print up some posters and maybe book inserts with it. Please reach out!

#Artists #3DArt #GraphicsDesign #ArtCommunity #Artwork

HOLY SHIT Y'ALL!

Scientists have found an obsidian tool making "workshop" which dates to 1.2 MILLION YEARS AGO in Ethiopia.

This pushes this type of tool making back 500,000 years.

WOOOOOOOOOOOW

https://www.vice.com/en/article/n7zx77/archaeologists-discover-12-million-year-old-workshop-in-mind-blowing-find

If you’re looking for inspiration on futuristic megastructures for your writing… you can’t go wrong looking through some work by #MarcelDeneuve #SciFi #WritingCommunity There’s a whole civilisation in this one image. #Brilliant #Art https://www.artstation.com/artwork/KrkQ0R

Pass Through Fire

#CriticalRole #TheLegendofVoxMachina #Keyleth #MastoArt #Art

#introduction #TwitterMigration
Hi folks! I'm a computer scientist who cares about equity and where tech meets society (#MachineLearning #AI #AlgorithmicFairness).

I currently do tech policy at the White House Office of Science and Technology Policy (#AIBillOfRights https://www.whitehouse.gov/ostp/ai-bill-of-rights/ )

I'm also a #ComputerScience #professor #researcher at Haverford College, co-founder #FAccT Conference, former Data & Society fellow.

Always happy to talk tech & society, #equity, #policy, and #teaching