James Hays (@jhhays) profile picture

Associate Professor at Georgia Tech and Director of Perception Research at Overland AI

Atlanta and Pittsburgh
Joined March 2012
Photo by James Hays (@jhhays)Photo by James Hays (@jhhays)Photo by James Hays (@jhhays)
Argoverse is hosting four competitions this spring: Motion Forecasting, Lidar Scene Flow, 4D Occupancy Forecasting, and End-to-end Forecasting. Top entries will be highlighted at the CVPR 2024 Workshop on Autonomous Driving. More details here: argoverse.org/tasks.html
3
6
1
32
6,767
James Hays retweeted
Can a model tell if a spoon hits a hardwood or carpet floor based on the sound? In our #ICCV2025 paper Clink! Chop! Thud!, we train an audio-vision-language representation that links the sound of an interaction with the objects that make it. (Sound on! 🎧)
2
2
8
538
Fascinating to be reminded that image manipulation dates back to the 1800s, especially in today’s generative AI era. Big thanks to @wjscheirer for an insightful talk at @mlatgt ! Check out his new book on the history of fake images and their social impact: sup.org/books/title/?id=3546…
4
17
1,261
We just hosted a Student Conference @mlatgt, bringing together students/fac from across campus to build a stronger community. Grateful to all participants & proud of the incredible work showcased! It was a success cuz we ran out of food :) Hear our Atlanta AI Buzz @GeorgiaTech!🐝
13
37
4,946
James Hays retweeted
I'll be presenting our work tomorrow (6/19) 5-630pm in Arch 4A-E (poster #329)! #CVPR2024
In our latest work, we introduce personalized residuals for efficient personalization of T2I diffusion models and localized attention-guided sampling that dynamically blends the learned concept into new contexts. #CVPR2024 cusuh.github.io/personalized…
1
3
23
3,747
James Hays retweeted
In our latest work, we introduce personalized residuals for efficient personalization of T2I diffusion models and localized attention-guided sampling that dynamically blends the learned concept into new contexts. #CVPR2024 cusuh.github.io/personalized…
2
5
1
22
6,277
Scene Flow promises to be a catch all solution for open-world detection of moving objects, but in practice they fail on the second most common class in AV datasets pedestrians! Solving this is the goal of this year's Argoverse 2 Scene Flow challenge! argoverse.org/sceneflow 🧵👇
1
7
1
14
3,960
Scene Flow methods promise to describe motion on any moving object… …but they can't even describe pedestrians walking! To focus the community on this problem, we're announcing the Argoverse 2 2024 Scene Flow Challenge focused on this issue: argoverse.org/sceneflow 🧵👇
1
4
6
789
Our work ZeroFlow: Scalable Scene Flow via Distillation is accepted at #ICLR2024! We show that scaled up distillation from a noisy teacher on unlabeled data outperforms human supervision to capture SotA! 🌎Website: vedder.io/zeroflow 📄Paper: arxiv.org/abs/2305.10424
1
5
1
18
3,061
James Hays retweeted
Thank you to my collaborators (@jhhays, Cynthia, Krishna, Zhifei, and Tobias)! @mlatgt @AdobeResearch
2
4
1,363
Argoverse is hosting 4 competitions this spring: Multi-agent Motion Forecasting eval.ai/web/challenges/chall… Unified Detection, Tracking, Forecasting eval.ai/web/challenges/chall… Scene Flow eval.ai/web/challenges/chall… 3D Occupancy forecasting eval.ai/web/challenges/chall… Please take a look!
1
3
2
25
5,557
Top performing methods will be highlighted at the CVPR 2023 Workshop on Autonomous Driving. cvpr2023.wad.vision/ #cvpr2023
4
657
Augmentations are important in ML, but did you know a soft augmentation is the way to go? Inspired by human visual classification studies, the more aggressive your image crop augmentations, the more your learning target softens. From my time at Argo! arxiv.org/pdf/2211.04625.pdf
2
3
20
2,946
James Hays retweeted
Argoverse 2: Next Generation Datasets for Self-Driving Perception and Forecasting abs: arxiv.org/abs/2301.00493
1
19
108
24,753
I continued exploring #stablediffusion's latent space over the weekend and oh my; there's still a LOT of treasure to be discovered inside this magnificent neural universe! Here's a quick thread with some of my personal favorites and how I found them..
18
231
42
1,796
Computing is the fastest growing College at Georgia Tech. So of course my home unit, the School of Interactive Computing, is hiring faculty. Stay tuned for a proper ad. For now, I polled our faculty and came up with this unofficial list of priority growth research areas.
2
18
1
99
James Hays retweeted
Working on a walkthrough of transformer code with side-by-side comparison to the computation graph. Felt the default image everyone uses is hard to interpret. Also, I like to show the matrices in my computation graphs. § refers to sections in a colab notebook (coming soon)
27
165
8
1,344
Replying to @jhhays
@jhhays is our guest lecturer at I2DL this semester! @TU_Muenchen Starting with the awesome Argoverse2 dataset! @argoai
3
19
James Hays retweeted
Excited to share "CoGS: Controllable Generation and Search from Sketch and Style" at #ECCV2022! (w/ Gemma Canet Tarrés, @tuvbui, @jhhays, @JCollomosse, Zhe Lin) Project page: cusuh.github.io/CoGS/
1
3
1
18
Some pictures of tonight's lunar eclipse. Taken with a cell phone mounted to an 80 mm refractor.
18