@CSProfKGD

#CS Associate Prof @YorkUniversity #ComputerVision Researcher, @VectorInst Affiliate @UofT status only, @ELLISforEurope Member, X-social media chair 4all vision

 Toronto, Ontario
csprofkgd.github.io
Joined August 2015
  • Tweets41.8K
  • Following202
  • Followers88K
6.7KMedia
I’ve launched 2 new course websites for my Introduction to #ComputerVision and #DeepLearning in Computer Vision courses. 🆕 Gaussian Splatting 🆕 Flow Matching 🆕 Sparse Autoencoders 🆕 Lecture Notes (work in progress) The videos don’t include voiceovers, but that’s coming in a future revision 🤞 Share your feedback in the thread.
8
31
1
324
44,896
Kosta Derpanis retweeted
Ideogram
@ideogram_ai1h
Today, we’re introducing Ideogram's first vector model. Generate clean, editable SVGs with smooth gradients and sharp typography from a text prompt, or turn any image into a detailed, scalable vector. Live in Ideogram and the API: ideogram.ai/models/vector
133
159
183
517
49,578
Kosta Derpanis retweeted
Vincent Sitzmann
@vincesitzmann14h
I have talked much about the "bitter lesson" of computer vision, arguing that many of the intermediate tasks that vision has solved traditionally - 3D reconstruction, segmentation, tracking, etc - are not necessary to build AI that understands the world through vision... (1/n)
Vincent Sitzmann
@vincesitzmannFeb 16
In my recent blog post, I argue that "vision" is only well-defined as part of perception-action loops, and that the conventional view of computer vision - mapping imagery to intermediate representations (3D, flow, segmentation...) is about to go away. vincentsitzmann.com/blog/bit…
5
17
2
222
19,953
Kosta Derpanis retweeted
Ulugbek S. Kamilov
@prof_kamilov6h
The new website of the Computational Imaging Group (CIG) at University of Wisconsin–Madison is now live. uw-cig.github.io
1
1
19
929
Kosta Derpanis retweeted
Xingjian Bai@SimulatedAnnealOct 21
1/9 Current optimizers are static during training. What if the optimizer could think and react while the network trains? We let an agent watch the training dynamics and rewrite the optimizer's source code as training goes: a reactive optimizer. In every language and image setting we tried that has no tuned recipe, it matches or beats the best recipe one can one-shot, and it guards against loss spikes and overfitting by itself. In the near future, optimization decisions may belong to agents, not humans. 📝 xingjianbai.com/reactive-opt… 💻 Try it on your task: github.com/xingjian-bai/opal 🧵
63
172
58
1,300
144,089
Kosta Derpanis retweeted
tom white
@dribnet19h
NeurIPS Creative Track 2026 is seeking sponsors to help fill high-visibility public areas of the NeurIPS convention center with art. If you or your company would like to help make this happen, please DM me this week!
🤖 Made with AI
2
3
46
201,463
Kosta Derpanis retweeted
Yi Ma
@YiMaTweets7h
The lectures of my new course "Principles of Deep Representation Learning" are now available at the Youtube site: youtube.com/playlist?list=PL… with new lectures continuing to be added there. I will open-source all the slides later in the textbook website: ma-lab-berkeley.github.io/de…

Principles of Deep Representation Learning | HKU DATA8014

Principles of Deep Representation Learning Fall 2026 · HKU DATA8014 Course Textbook (Open Access): https://ma-lab-berkeley.github.io/deep-representation-lear...

youtube.com
6
67
2
513
26,927
Kosta Derpanis retweeted
Marques Brownlee
@MKBHDOct 21
The current state of apps on iPhone Duo: youtu.be/ESABTN4VtrA
1,017
1,566
1,040
45,809
9,308,436
Kosta Derpanis retweeted
Ashkan Mirzaei@ashmrz1022h
Check out Cosmos Simulation: Interactive World Simulation with Cosmos for Physical AI: 🌐Website: research.nvidia.com/labs/cos… 📜Report: research.nvidia.com/labs/cos… 🤖 Models: huggingface.co/collections/n…

Interactive World Simulation with Cosmos for Physical AI

Interactive world simulation with Cosmos: real-time, streaming models for camera navigation, robot manipulation, and autonomous driving.

research.nvidia.com
8
40
2,038
Kosta Derpanis retweeted
Kosta Derpanis
@CSProfKGDOct 21
If just reading X, it would appears a large number of #ICLR02027 submissions are AI slop 🤔 Anyone want to provide some balance to this narrative?
37
2
2
69
16,782
Kosta Derpanis retweeted
Sugam05
@SugamPanthi21h
Replying to @CSProfKGD
i'm packing it up to the farms, cya in 3 years!
1
3
720
😅
Lucas Beyer (bl16)
@giffmana21h
My brother made a fun website: deprecation.org Enter your job and it tells you when you'll be deprecated, and exactly how. For me it's... pretty soon and pretty accurate! He's a very creative person (artist turned frontend dev to pay bills, now obv worried by AI) and this is full of little nuggets and easter eggs. It's an eerie mix of fun and jokes and real worries. Check it out!
3
40
10,809
"Go with the flow" - Flow matching trailer next? 🎬 🤔
11
2,564
Today’s campus walk and read 🚶
1
5
61
6,037
Cool work! The observation that the effective receptive field in diffusion transformers (DiTs) is significantly smaller than the theoretical receptive field during the later stages of denoising reminds me of an earlier finding in convolutional networks. arxiv.org/abs/1701.04128
4
494
If just reading X, it would appears a large number of #ICLR02027 submissions are AI slop 🤔 Anyone want to provide some balance to this narrative?
37
2
2
69
16,782
#ICLR2027
1,546
Halfway through my computer vision course. Midterm day! 😅
1
21
2,571
Kosta Derpanis retweeted
Nenad Tomasev
@weballergyOct 21
I am happy to announce an opening for a student researcher on my team at GDM for 2027, based in London - for a project at the frontier of multi-agent coordination, in duration of 3 months with a possibility of an extension.
5
22
3
205
14,050
Kosta Derpanis retweeted
Kosta Derpanis
@CSProfKGDOct 19
How did I make the “Gaussians Unite” trailer? Concept → Story → Storyboard → Script → Codex + Blender + After Effects → Iterate. Iterate. Iterate.
Kosta Derpanis
@CSProfKGDOct 18
“Gaussians Unite” 🫟 Trailer for my Gaussian Splatting lecture, coming this winter to my #DeepLearning course 🎬
2
1
1
23
3,114
Kosta Derpanis retweeted
Grant Sanderson@3blue1brownOct 20
Replying to @ONagel33303
Oh, to be clear, I will very happily keep making videos long past when LLMs can generate clear explainers. I hope people have found value in my videos because they explain things clearly, but that's not even the highest-order bit for what makes a video land, in my opinion.
53
54
11
2,287
92,339
Load more