Subhankar Roy (@sroy907) header
Subhankar Roy (@sroy907) profile picture

Tenure track Assistant Professor at University of Bergamo.

Joined April 2010
Photo by Subhankar Roy (@sroy907)Photo by Subhankar Roy (@sroy907)Photo by Subhankar Roy (@sroy907)Photo by Subhankar Roy (@sroy907)Photo by Subhankar Roy (@sroy907)Photo by Subhankar Roy (@sroy907)Photo by Subhankar Roy (@sroy907)Photo by Subhankar Roy (@sroy907)Photo by Subhankar Roy (@sroy907)Photo by Subhankar Roy (@sroy907)
Man, what a day
39
894
73
8,730
275,626
Pulled another all-nighter for #CVPR. Heart says yes. Mind says no. Body says NOPE!
1
5
470
Subhankar Roy retweeted
I once tried to explain Bayesian reasoning at a dinner party. By dessert, everyone had updated their beliefs about inviting me again.
33
248
15
3,478
66,387
Subhankar Roy retweeted
Harry Kane going from football’s most famous trophyless player to a Ballon d’Or winner might be the greatest example of why changing employers can improve your life.
297
21,305
539
140,787
1,612,305
Paper got rejected from #WACV. Happens, no big deal. But what are the odds that every "reviewer" raises the exact same weakness as their very first bullet point in the weaknesses section? For context, the weakness is from the 5th page of the manuscript. EMBARRASSING!!!
6
555
Grieving my first major grant proposal rejection with my first ever sub-30 5K run. Let's call it a tie!
9
223
What is more disappointing? Having a paper rejected from #NeurIPS or not being able to present a paper which you co-authored due to space constraints?
1
11
1,980
Subhankar Roy retweeted
I'm currently writing and editing ICLR submission (and yes, I'm a dinosaur. I still do this by hand, no AI, just a spell checker) and I realized that when I work on a paper, at every step I try to imagine how someone would read this. Is the information clear? Will they get the point? Do they need more information (or less)? This is how communication works ...
14
18
2
440
27,199
Subhankar Roy retweeted
Seems like this idea is starting to get traction (e.g. ICLR). But limits like 20 papers per author are still ridiculous. Let's make it 3--5.
My (probably controversial) idea to improve the state of publishing/reviewing in computer vision's overloaded main conferences: a limit (e.g. 3) on the number of papers that any individual can submit as co-author to one conference.
3
3
63
8,865
Subhankar Roy retweeted
PPO: rejected from NIPS 2017
39
198
44
2,592
534,162
Subhankar Roy retweeted
Two fully-funded PhDs at CRIL (CNRS / Université d'Artois, Lens co-supervised by me. Very different topics
2
6
1
19
3,504
Subhankar Roy retweeted
The variance across ICLR scores is wild, OpenAI should decrease the temperature
8
44
7
608
34,614
Subhankar Roy retweeted
🌺 Aloha! I will be presenting our work, "FedMVP: Federated Multimodal Visual Prompt Tuning for Vision-Language Models" at #ICCV2025 in Hawaii 🇺🇸 ! 🗓️ Wednesday 22 October, Exhibit Hall I, Poster #1650, Session 4, 14:30 - 16:30 (HST) @eliricci_ , @sroy907 , @BiplabB81293285
1
3
136
Subhankar Roy retweeted
Replying to @GFaang97609
no i never run ablations, because i believe god has a plan
8
19
1
190
14,853
Having a paper, which was accepted by the AC, rejected by the PCs is really atrocious. This can damage the morale of young researchers, especially PhD students. Quota-based acceptance should be stated in the cfp so that the authors can calibrate their expectations. #neurips25
1
6
572
We submitted a paper on semantic segmentation to a CV conference that shall not be named. Reviewer's comment, "The approach involves extracting high-quality domain knowledge from expert-written documents" Dear reviewer, kindly choose a better LLM next time.
6
251
This year 11 researchers from University of Oxford have received ERC starting grants! To put things into perspective to those who don't know, an institute is fortunate to have just one accepted. Success rate is ~12%. Congratulations to all the recipients.
1
1
156
The best research teams are not the ones with the most people, but the ones with the right people.
12
8
213
23,269
The "O" in OpenAI stands for Obviously open-source. Obviously as in Alan Rickman obviously.
The “O” in OpenAI stands for open-source.
1
197