Researcher at @Google, 3D computer vision & machine learning. Previously PhD at ETH Zurich, intern at @Google, @Meta, @Microsoft, @magicleap.
Zurich, Switzerland
- Tweets 428
- Following 479
- Followers 6,978
- Likes 1,810
Pinned Tweet
Introducing GeoCalib for camera calibration: gravity & intrinsics from a single image 📸
➡️Differentiable geometry optimization FTW!
➡️Video piped.video/uOmTwvKreM4
➡️Demo veichta-geocalib.hf.space
➡️Paper arxiv.org/pdf/2409.06704
by @veichta with @PhilippCSE @mapo1 for #ECCV2024 1/
6
30
242
32,787
📢 Code release: a new toolbox for visual localization
github.com/cvg/Hierarchical-…
➡️ try out SuperGlue for localization - CVPR'20 winner
➡️ run it with SfM on your own dataset within minutes
➡️ evaluate your local features/retrieval
➡️ take part in ECCV'20 challenges: 3 weeks to go!
3
137
545
Next week at #CVPR2020: our oral paper "SuperGlue: Learning Feature Matching with Graph Neural Networks" with @ddetone, @quantombone, A Rabinovich
Website: psarlin.com/superglue
Paper: arxiv.org/abs/1911.11763
Video: piped.video/watch?v=BNaIGI4V…
Winner of several CVPR competitions 👇🏻
5
96
436
You liked SuperGlue? You'll love ⚡️LightGlue⚡️, our new deep network for light-speed image matching!
➡️Faster, stronger, easier to train than SuperGlue
➡️Code: github.com/cvg/LightGlue
➡️Paper: arxiv.org/pdf/2306.13643.pdf
Fantastic work by @PhilippCSE for #ICCV2023, with @mapo1
1/
6
83
424
42,995
📢 Introducing LaMAR: a benchmark for Localization and Mapping with AR devices
Website: lamar.ethz.ch
Video: piped.video/32XsRli2coo
➡️ Join our LaMAR tutorial at #ECCV2022 on Monday 24/10 - intro by Marc Pollefeys @mapo1
⬇️ hours of AR data, 3 locations, over 2 years
9
95
366
📢 We just released the code for OrienterNet, a deep net that learns to localize in 2D maps 🗺 like OpenStreetMap, just like humans do! Join us today at #CVPR2023 poster PM-98.
Video: piped.video/wglW8jnupSs
Code: github.com/facebookresearch/…
Try our demo: take a picture and localize!
4
101
340
54,736
I defended my PhD thesis last week 🥳 Thank you to everyone that made this possible, including my advisor @mapo1, examiners @Jimantha @quantombone and Daniel Cremers, and the amazing @cvg_ethz. As per the tradition, I received a nice commemorative hat 🎓 Now time for vacations 😎
30
7
346
21,725
What makes a good @CVPR poster?
➡️ lots of visuals to guide your presentation
➡️ thick paper: serves as comfy nap mat 😴
➡️ glossy paper: serves as easy-to-clean tablecloth 🍽
➡️ little text: serves as bright windshield sunshade ☀️ 🚗
Tested & 💯 approved for camping in Canada!
8
24
316
26,225
Career update: after a fun post-PhD summer break, I'm excited to join Google Zurich as a researcher 🤩 I can't wait to push the limits of world-scale mapping at Geo (Maps, StreetView, Earth) with a stellar team and to contribute to Maps, one of my favorite products 🚀
11
1
282
12,399
📢 Coming at #CVPR2021: our paper "Back to the Feature: Learning Robust Camera Localization from Pixels to Pose" w/ @aaunagar @HugoGermain17 @mapo1 @SattlerTorsten etal
Website: psarlin.com/pixloc
Paper: arxiv.org/abs/2103.09213
Video: piped.video/vPkXhKQn2oI
Cool visuals 👇
6
43
208
New release of our toolbox hloc: run 3D reconstruction and visual localization with COLMAP in Google Colab!
➡️ SfM, painless, with GPUs, for free
➡️ GitHub github.com/cvg/Hierarchical-…
➡️ Colab colab.research.google.com/dr…
Thanks @PhilippCSE @j___lambert @mihaidusmanu for the awesome work!
1
45
171
For #NeurIPS2023 we're presenting our latest work SNAP - Self-Supervised Neural Maps for Visual Positioning and Semantic Understanding
Geo data at scale + self-supervision = 💥
Video: piped.video/xogrwdgNQdo
Paper: arxiv.org/pdf/2306.05407
Code: github.com/google-research/s…
summary🧵⬇️
ALT Teaser image
4
42
160
22,355
Yesterday the #ICCV2023 workshop Quo Vadis Computer Vision was full of takes so hot I needed to let them cool down for a while not to burn myself - my favorites ⬇️
> We have a fanatic adherence to experiment, mostly hilariously poorly conducted e.g. no error bars – Forsyth
1/
3
18
154
46,688
📢Today we release the @ETH-@Microsoft visual localization dataset as part of the visloc #ICCV2021 challenge! Focus on AR in changing conditions: in/outdoor, day/night, @HoloLens & phones
➡️github.com/cvg/visloc-iccv20…
Demo with hloc & SuperGlue: setup in 5 minutes or your money back!
3
32
145
🚨 This week at #ICCV2021, check out "Pixel-Perfect Structure-from-Motion with Featuremetric Refinement" w/ @PhilippCSE V. Larsson @mapo1
➡️ oral & *best student paper award*
Website: psarlin.com/pixsfm
Paper: arxiv.org/abs/2108.08291
Video: piped.video/2HuCMuraFk0
thread ⬇️
7
23
147
I won the best presentation award at #ICVSS2022 🥳 Many thanks to the school committee and to all attendees for voting! I very much enjoyed presenting and chatting in-person after 2 years of virtual posters sessions 😊
10
7
145
We have released COLMAP implementations of recent papers that make Structure-from-Motion robust to symmetries
Surprise: none of these SOTA algorithms works consistently across datasets with the same hyperparams. More work needed for this difficult problem!
github.com/cvg/sfm-disambigu…
2
22
121
Yesterday at #ICVSS2022: incredible talk by @vincesitzmann on neural scene representations, organizing many recent works into a unified taxonomy with great clarity.
2
3
105
Wow so much exciting progress on Structure-from-Motion this week! Congratulations to @jianyuan_wang @JeromeRevaud @eric_brachmann @vincesitzmann & teams for their fantastic works! Now we really feel pressured to improve COLMAP before everyone catches up 😉
4
3
96
12,436
Personal update: this week I joined the Microsoft MR&AI Zurich lab as a research intern :) I am excited to work on cool 3D geometry+learning problems with @mapo1 and other inspiring researchers!
6
97
📢Join us tomorrow Friday morning at our #ICCV2023 poster S-6 to learn more about LightGlue!
💥We have released the training code at github.com/cvg/glue-factory
➡️You can now train LightGlue with SIFT, ALIKED, or your own local features, on your own dataset
⏩More models coming soon!
You liked SuperGlue? You'll love ⚡️LightGlue⚡️, our new deep network for light-speed image matching!
➡️Faster, stronger, easier to train than SuperGlue
➡️Code: github.com/cvg/LightGlue
➡️Paper: arxiv.org/pdf/2306.13643.pdf
Fantastic work by @PhilippCSE for #ICCV2023, with @mapo1
1/
2
11
100
11,776
Very cool work by Frank Fu & @MauriceFallon extending Back to the Feature to Camera-Lidar calibration. Learned image-lidar representations + differentiable geometry optimization FTW!
As expected, great generalization compared to black-box deep nets.
📢 NEW PAPER: Self-Supervised Camera/Lidar Calibration.
Appearing at @corl_conf '23: We've developed a self-supervised method for camera/LiDAR correspondence learning unlocking long-term sensor fusion in the wild.
2
10
90
17,547
Really sad to miss #CVPR2024 and its fantastic papers & workshops! But Puglia/Italy is treating me well in return 😊 I hope to catch up with everyone at ECCV in Milan - enjoy Seattle!
2
94
8,756
We have now released the full raw data of LaMAR, which include 3D lidar point clouds & meshes, on-device ToF depth maps, IMU data, GPS signals, etc. Huge potential to benchmark and train models for different geometric vision tasks!
Check out lamar.ethz.ch to gain access
📢 Introducing LaMAR: a benchmark for Localization and Mapping with AR devices
Website: lamar.ethz.ch
Video: piped.video/32XsRli2coo
➡️ Join our LaMAR tutorial at #ECCV2022 on Monday 24/10 - intro by Marc Pollefeys @mapo1
⬇️ hours of AR data, 3 locations, over 2 years
2
24
88
15,756
We have finally released the code for our #ICCV2021 Pixel-Perfect Structure-from-Motion 🥳 Easy to use with COLMAP and hloc, more accurate/optimized/scalable than our original research code, can fit many use cases. Kudos to @PhilippCSE for the nice work
➡️ github.com/cvg/pixel-perfect…
🚨 This week at #ICCV2021, check out "Pixel-Perfect Structure-from-Motion with Featuremetric Refinement" w/ @PhilippCSE V. Larsson @mapo1
➡️ oral & *best student paper award*
Website: psarlin.com/pixsfm
Paper: arxiv.org/abs/2108.08291
Video: piped.video/2HuCMuraFk0
thread ⬇️
1
13
84
Quick update: we have finally released the code for our CVPR 2021 paper Back to the Feature: github.com/cvg/pixloc It includes the evaluation & training as well as a 3D viewer, a dataset downloader, etc. ↘️
📢 Coming at #CVPR2021: our paper "Back to the Feature: Learning Robust Camera Localization from Pixels to Pose" w/ @aaunagar @HugoGermain17 @mapo1 @SattlerTorsten etal
Website: psarlin.com/pixloc
Paper: arxiv.org/abs/2103.09213
Video: piped.video/vPkXhKQn2oI
Cool visuals 👇
1
11
82
Yesterday was my last day interning at Facebook/Meta @RealityLabs. This was an incredible experience: I had fun working on cool problems with amazing people, from which I learned so much! Now I can't wait to soon tell you all about what I worked on 🔜💥
73
I'm excited to attend @eccvconf this week in Milano!
➡️ I'll give a short talk at the Project Aria tutorial on Sunday: projectaria.com/events/eccv2…
➡️ Join the GeoCalib poster on Wednesday 10/02 16:30-18:30 #327 with live demo!
➡️ Feel free to reach out to meet & chat 😀
#ECCV2024
Introducing GeoCalib for camera calibration: gravity & intrinsics from a single image 📸
➡️Differentiable geometry optimization FTW!
➡️Video piped.video/uOmTwvKreM4
➡️Demo veichta-geocalib.hf.space
➡️Paper arxiv.org/pdf/2409.06704
by @veichta with @PhilippCSE @mapo1 for #ECCV2024 1/
1
6
74
7,282
This would have not been possible without the amazing @ddetone @quantombone and the support from @rsiegwart 🥳
Paul-Edouard Sarlin is awarded the Graduate Award 2020 of the SEW-Eurodrive Foundation for his Master's thesis on "SuperGlue: Learning Feature Matching with Graph Neural Networks", supervised by Professor Roland Siegwart. Congratulations! @pesarlin
7
63
I am excited to share that I joined @mapo1 Prof Marc Pollefeys' CVG group at ETH as a PhD student! The next years will be a lot of fun & challenges, working on 3D vision & learning, with awesome colleagues & in the great city of Zurich 😁🇨🇭
6
1
64
On July 13th, I'll give a talk on how deep learning can supercharge 3D reconstruction and world mapping. I feel very lucky to present next to stellar world-class teams from Meta, Google, Microsoft, Pix4D. Join us if you're in Zurich!
Join us on July 13th at Meta Zurich for the 3rd Computer Vision Zurich Meetup on "Building a Digital World". Speakers from academia & industry will discuss about 3D reconstruction & mapping.
Seats are limited!
meetup.com/computervisionzur…
w/: @FedericaBogo,Margarita Grinvald,@fedassa
1
6
64
9,825
My hloc toolbox with SuperGlue won 4 visual localization challenges at #ECCV20 🔥 I will give two talks next week at two workshops
➡️for Autonomous Driving, Sun 23rd sites.google.com/view/mlad-e…
➡️for Long-Term Changing Conditions, Fri 28th sites.google.com/view/ltvl20…
See you "there" 🤩
📢 Code release: a new toolbox for visual localization
github.com/cvg/Hierarchical-…
➡️ try out SuperGlue for localization - CVPR'20 winner
➡️ run it with SfM on your own dataset within minutes
➡️ evaluate your local features/retrieval
➡️ take part in ECCV'20 challenges: 3 weeks to go!
1
14
57
I will give a total of 4 talks during #CVPR2020:
- @ 3D Scene Understanding workshop scene-understanding.com/sche… Monday
- SuperGlue for localization sites.google.com/view/visloc… 2x talks Monday
- SuperGlue for two-view+SfM image-matching-workshop.gith… Friday 11:45a PST
+ live QA Wed 10:40a PST
Next week at #CVPR2020: our oral paper "SuperGlue: Learning Feature Matching with Graph Neural Networks" with @ddetone, @quantombone, A Rabinovich
Website: psarlin.com/superglue
Paper: arxiv.org/abs/1911.11763
Video: piped.video/watch?v=BNaIGI4V…
Winner of several CVPR competitions 👇🏻
2
7
58
Our hierarchical localization paper won the visual localization challenge at #CVPR2019 😀 Come to my talk at the visual localization workshop on Monday 2:15PM, and to our poster 3.2/213 on Thursday!
Paper: arxiv.org/abs/1812.03506
@CVPR2019 @ETH_en @ASL_ETHZ @7Srobotics
1
16
53
What's the CO2 footprint of a conference like @CVPR? Including 1) the energy used by all GPUs for all experiments of all papers, and 2) flights for all in-person attendees. Has anyone ever looked into this? and for other academic conferences?
9
3
51
13,141
Just landed in Sicily to attend #ICVSS next week! This is my first in-person research event since 2019 - super excited to finally meet the real faces behind the names, make new friends, and learn from everyone! Let's meet if you're attending too 😀
1
1
53
Today I presented my Master thesis "SuperGlue: Learning Feature Matching with Graph Neural Networks" arxiv.org/abs/1911.11763
done with @ddetone @quantombone @ASL_ETHZ @ETH_en
This marks the end of my time at Magic Leap - it was an amazing experience! 😁 Stay tuned for what's next
9
48
Are you attending #CVPR2023 @CVPR and interested in image matching? Then you should attend the Image Matching Workshop tomorrow morning! image-matching-workshop.gith… Bonus: team RMD-3DV, 8th place at the Kaggle challenge, will preview a pretty cool work in their talk at 12:15 😉
1
7
48
10,364
What a stunning view for @3DVconf in Davos! It’s been very fun meeting so many people doing great work in the space of 3D vision 😀
46
3,884
SuperGlue won 2 visual localization challenges on local features & indoor/outdoor loc ➡️ check out the workshop sites.google.com/view/visloc…
It is also ranks 1st in the image matching challenge of @ducha_aiki @kwangmoo_yi @FuaPv by a large margin ➡️ leaderboard vision.uvic.ca/image-matchin…
2
6
41
#ECCV2022 was so fun! Great discussions, delicious food, wild parties, and many new friends. Finally got to meet the humans behind the names & papers - they’re indeed all real 😀
1
1
41
Our benchmark paper 𝗙𝗶𝘀𝗵𝘆𝘀𝗰𝗮𝗽𝗲𝘀 got accepted as an oral to the #ICCV19 workshop on Autonomous Driving! 🙂 Join on Monday if you're interested in out-of-distribution detection for semantic segmentation!
Paper: arxiv.org/abs/1904.03215
Benchmark: fishyscapes.com/
1
6
40
"We must take the risks of AI as seriously as climate change. We can’t afford the same delay with AI."
This planet will be hell long before AI becomes smart enough to be an existential risk: ~30% of the population lives in areas unliveable by 2070. AI risk is a mere distraction.
AI risk must be treated as seriously as climate crisis, says Google DeepMind chief theguardian.com/technology/2…
2
4
39
8,081
Hey Computer Vision Twitter,
I will be at TU Munich @TU_Muenchen on Friday this week - feel free to reach out if you want to grab coffee and chat :)
1
1
41
tldr: PixLoc = end-to-end camera localization by learning deep features & uncertainties 😎 The code will be released in the next days! In the meantime:
➡️ Join our poster: Tues 6:8.30am CET
➡️ Play with our interactive viewer at psarlin.com/pixloc
More viz coming soon 🤩
2
5
36
This will have a long-lasting impact on their career. Attending virtually is just pretty bad for networking. How many lost the chance to meet their future PhD/postdoc advisor? Start fruitful collaborations? Learn this new unwritten trick at poster xyz? Please never again. 2/2
36
2,703
Attending #ECCV2022? Join us today 24/10 for our LaMAR Tutorial 14:00-17:00 in room G&N - everything you need to know about the dataset, the benchmark, and why we built it! @eccvconf @mihaidusmanu @visionviktor @mapo1
📢 Introducing LaMAR: a benchmark for Localization and Mapping with AR devices
Website: lamar.ethz.ch
Video: piped.video/32XsRli2coo
➡️ Join our LaMAR tutorial at #ECCV2022 on Monday 24/10 - intro by Marc Pollefeys @mapo1
⬇️ hours of AR data, 3 locations, over 2 years
2
8
34
LightGlue is already in @kornia_foss thanks to @ducha_aiki! It is also much easier to train - only 5 GPU-days. We hope that this will make it much easier to do research on image matching. We will soon release the full training code under Apache 2.0 license - STAY TUNED ⏩
3/3
3
6
34
4,250
I'll be in London this week - anyone in the Computer Vision / Robotics space interested in meeting up for a chat? 🙂
1
31
It was great to finally present a poster in person and chat research with wine! Looking tired after presenting for 4 hours 😅
1
32
I finally got to try the famous giant slides at @TU_Muenchen and had so much fun giving a talk at Daniel Cremers’ group. Thank you @felixwimbauer @JennySeidensch1 @12Zuo @KoestlerLukas and others for the nice invite and warm welcome - with beers and cake!
1
30
This. By writing many weak papers, everyone in our community is loosing out:
1) it overburdens the review system - reviewers handle too many papers and thus don’t have time to give useful feedback
2) it adds noise to arxiv/conferences and makes it easier to miss great papers
⬇️
Since we just wrapped up an AI megaconference, it felt like a good day to plead for fewer papers. argmin.net/p/too-much-inform…
2
1
28
9,101
#ICCV2021 nominated me as outstanding reviewer 🥳I enjoyed the time spent reviewing & learned a lot from the discussions with the ACs and other reviewers - thank you very much! It's a great way to learn about common issues in submissions and improve one's paper writing!
[PCs Update] We acknowledge 210 outstanding reviewers (top 5% experienced and top 5% student reviewers) online at:
iccv2021.thecvf.com/outstand…
amongst the many amazing reviewers this year.
We also ack. generous emergency reviewers:
iccv2021.thecvf.com/emergenc…
25
Good question! ResNet has higher capacity + memory-efficiency but features not well-localized vs VGG because: 1) aggressive downsampling by 1st conv, 2) downsampling with strided conv instead of max-pool. This is fixable but with some network surgery, so VGG is a simpler choice
3
1
27
This is a great read! I've also been thinking about this for a while: books (and the internet) are a summary of human knowledge; language models, even if mere approximations of it, encode statistics of the world useful to solve many vision and robotics tasks. thoughts ⬇️ 1/n
Here is the sequel to "Just ask for Generalization" - in this blog post I argue that Generalization *is* Language, and suggest how we might be able to re-use Language Models as "generalization modules" for non-NLP domains. Check it out!
evjang.com/2021/12/17/lang-g…
2
2
26
This paper is a great example of deliberate misleading evaluation of baselines. Why? ⬇️
1/ To localize with local features we need a 3D sparse map. How it's done in the paper: lift 2D keypoints to 3D using depth from OpenCV stereo block matcher. Not joking.
LM-Reloc: Levenberg-Marquardt Based Direct Visual Relocalization arxiv.org/abs/2010.06323 #computervision
7
2
27
The #ECCV2022 LaMAR tutorial yesterday was a success, with interesting discussions & feedback - thanks everyone for joining! The slides are now online: lamar.ethz.ch/tutorial-eccv2…
@eccvconf @mapo1 @visionviktor @mihaidusmanu
Attending #ECCV2022? Join us today 24/10 for our LaMAR Tutorial 14:00-17:00 in room G&N - everything you need to know about the dataset, the benchmark, and why we built it! @eccvconf @mihaidusmanu @visionviktor @mapo1
6
26
My take: folks working on AI know that any AI existential risk is decades away and that we're not much closer to it now than pre-deeplearning. Those that signed the statement either seek attention (=funding) or want to block competitors from entering the juicy chatbot market. 1/
Do you see a value in engaging in these discussions? Myself I do not participate but sometimes I'm worried these folks will end up overregulating the field and causing damage that could be avoided if I (we) spoke out more loudly. What's your take?
2
2
24
6,218
Alternatively: because similar papers are grouped in the same poster session? Authors that are interested in one paper are busy presenting their own poster at the same time! This time conflict indeed made me miss several great papers today.
1
22
> 99% of the old literature was crap, but so is 99% of the current one – Forsyth
> Robotics is too important to be left to roboticists – Malik
3/
ALT The mood among Computer Vision researchers
3
1
22
29,567
I’m attending #ICCV2033 this week in Paris 🇫🇷 Looking forward to meeting old and new friends - please reach out if you have time for a chat! This year I’m super excited about the great line up of workshop speakers 🤩 @ICCVConference
1
21
3,601
Hey @GMFarinella, where are you? ICVSS continues in Syracuse! @barbafrank @AdrianoSimonet4 @ElenaCamuffo @Luna31683093 @felixwimbauer @sophiajwagner
4
21
My thread yesterday could be seen as offensive and sparked concerns among many. Let me rephrase ⬇️
1/ My goal was not to publicly attack the authors of this paper, for which I have immense respect and sympathy, but instead to bring awareness on an incorrect evaluation protocol.
This paper is a great example of deliberate misleading evaluation of baselines. Why? ⬇️
1/ To localize with local features we need a 3D sparse map. How it's done in the paper: lift 2D keypoints to 3D using depth from OpenCV stereo block matcher. Not joking.
1
16
We’re excited too, @cvg_ethz FTW! There will also be a workshop dedicated to SfM 50sfm.fbk.eu
3
19
2,812
Needless to say: this was one of the most fun and insightful workshop I’ve ever attended. Much more food for thought than technical workshops. Thank you @georgiagkioxari @sainingxie @mia_chiquier @TongPetersb for organizing it!
Yesterday the #ICCV2023 workshop Quo Vadis Computer Vision was full of takes so hot I needed to let them cool down for a while not to burn myself - my favorites ⬇️
> We have a fanatic adherence to experiment, mostly hilariously poorly conducted e.g. no error bars – Forsyth
1/
1
1
16
6,769
It's great to have so much exposure just 4 months through my PhD. Big thanks to my past & current co-authors & supervisors, from which I learn(ed) so much: @quantombone @ddetone @mapo @JuanNie17024970 @marcin_dymczyk @ASL_ETHZ @rsiegwart and many others
16
🏎💨LightGlue is fast thanks to a mechanism that reduces the depth (# of layers) and width (# of points) *adaptively* for each input & layer. This means that easy image pairs are matched much faster than difficult ones - great for applications like SLAM & Structure-from-Motion
2/
1
3
18
2,542
I disagree with "ML research is not about visualizations". Doing research requires understanding: we cannot explain ideas to others without first understanding them fully. ML theory alone is often not enough so visualizations are critical - good scientists need these skills.
1
1
17
At the #CVPR2023 CIVILS workshop, @quantombone is giving an overview of OrienterNet, our new work on localizing with OpenStreetMap - stay tuned for more details or come to our poster on Thursday afternoon!
1
2
16
1,438
PixLoc learns which features and objects matter for robust, long-term localization - a small step towards spatiotemporal understanding of appearance changes for human-level localization! 🤖
2
16
- Benchmark: visuallocalization.net/bench… (look for "Hierarchical Localization")
- ECCV'20 challenges: sites.google.com/view/ltvl20… organized by @SattlerTorsten @mapo1 @vbalnt @hulahuub & others
- Hierarchical localization paper: arxiv.org/abs/1812.03506
2
15
3) most importantly, it distracts us from writing great papers, which requires sustained, focused effort. Great papers have the highest impact on your career, while having a paper accepted to a conference is not special anymore - the review process is not sufficiently selective…
1
14
2,516
> We are experimentalists but we should feel lucky: unlike Monkeys, GPUs don’t need food, they just do the work.
> The problem with Putin, Musk, etc. is that they lack feedback from the real world – stay connected!
– Efros
2/
ALT Checking in with the real world
2
11
2,451
The audience asking for clarification: “I’m not so familiar with Star Wars, but aren’t stormtroopers the bad guys? Then how can training to become a good Jedi prepare us to later be a stormtrooper?”
= industry vs academia, competing or complementary?
4/4
ALT Industry vs Academia
4
15
3,280
The review process might be random but whether a paper has impact or not is much less random - citations/twitter decide so, not reviewers. Writing great papers takes sustained, focused effort and is harder if you get distracted by smaller, low-impact projects along the way.
1
2
14
2,841
#icvss2022 was incredible! Thank you @GMFarinella and the whole organizing team for your dedication, energy, and time! You're creating lifelasting memories and powerful positive reinforcement for generations of researchers. The computer vision community is truly amazing 👏
Last Social @ ICVSS 2022
Yes, ICVSSers are young and they have to go down with their “mind stress” after a week with 40 hours of scientific activities!
I already miss 2022 ICVSS Students! They will depart tomorrow morning ;-(
13
Then @MattNiessner told us about 3D reconstruction. I love this conclusion: combine deep nets, to learn priors and compressed representations, with geometry to enforce problem constraints. Use learning with parcimony!
1
13
Good luck running COLMAP on such dense matches - it will take forever. But hang on a little longer, we have something for this 😉
3
13
1,175
The result: hundreds of AR unconstrained trajectories, aligned with ground-truth poses & scene geometry - perfect not just for localization & mapping but also for many other geometric vision tasks!
1
1
12
OrienterNet uses global & free maps from OpenStreetMap so you can try the demo in *any* city in the world!
Work done during an internship at @Meta @RealityLabs with @ddetone, @shamangary, A Avetisyan, @jstraub6, @quantombone, S Rota Bulo, @rapideRobot, P Kontschieder, V Balntas
2
2
14
2,386
Nice work!
1) How much faster is it than our Python code?
2) Do you want to help us integrate it into COLMAP?
3) Did you profile SfM before trying to move it to GPU? COLMAP's BA can run on GPU via Ceres/PBA. The remaining ops are already fast on CPU and not so GPU-friendly.
2
14
2,177
Yes, I should have made it clear - I would certainly change the default retrieval in hloc (and soon COLMAP!) to any approach that provides clear benefits over NetVLAD across the board :)
1
1
12
1,100
2/ The correct way: triangulation+BA. Every CV student learns that. Yet the paper deems this "too slow" vs an extra-engineered DSO. I already called this out in my talk in August piped.video/M-X6HX1JxYk?t=5245 + it is really so easy & fast with hloc+COLMAP github.com/cvg/Hierarchical-…
2
3
12
Plugging in some experiments we did last year: none of the heuristics of the recent literature works consistently - it's just a hard problem. Still waiting for a deep net to solve it (doable but getting enough training data is hard)
We have released COLMAP implementations of recent papers that make Structure-from-Motion robust to symmetries
Surprise: none of these SOTA algorithms works consistently across datasets with the same hyperparams. More work needed for this difficult problem!
github.com/cvg/sfm-disambigu…
1
2
12
+ images have perfect GT poses
- queries are synthetic: occlusion artifacts, unrealistic lighting
- only 1 scene (3 in LaMAR)
- much fewer images (x10 in LaMAR)
- reference images from laser scanner with large FoV: easier to localize with, but too sparse to triangulate 3D points
1
4
11
2,062
On top of the dataset, we also present:
2️⃣ a pipeline to build such ground truth fully automatically - no manual effort!
3️⃣ an extensive analysis of the state of the art, with great insights about using additional sensors like wifi/bluetooth, image sequences, VIO poses!
1
1
12
"Could you please send us the paper that convincingly shows that a dumb rock has 0% chance of causing human extinction?" You're not doing science if your theory is not falsifiable.
To make that happen, could you please send us the paper that convincingly shows that human-level AI has a 0% chance of leading to human extinction? That the four independent ways that e.g. Dan Hendrycks found (arxiv.org/abs/2306.12001) that could lead to extinction are all wrong?
1
1
10
4,159
Really cool, thanks for sharing! We also had a lot of fun visualizing attention patterns for SuperGlue, Fig 16 arxiv.org/abs/1911.11763 - highlights the importance of non-locality even for low-level vision tasks.
1
9
2/ I think that this paper is of great value to the community and I am glad that it got accepted to #3DV2020. But I am also a proponent of "good" science and want to discourage future works from adopting the evaluation as is, which is a common pattern in Computer Vision.
1
7
They are all evaluated only on outdoor datasets, right? I would love to see some evaluations with indoor data, for example on LaMAR, where retrieval is the main bottleneck lamar.ethz.ch/
2
1
8
2,313