Siyu Tang @VLG-ETHZ @SiyuTang3 Twitter profile

Last Seen Profiles

@iriina333

@tad55kawasaki

@50x48x48

@Aracash

@tadataka_k

@daniellouisy

@lageloni

@lowbatreyy

@Infxnii

@SCA_Hoops

@chartslizzo

@bankmansneed

@shirohebibocchi

@chhai22

@blsdsyd26456

@schadenfreude74

@the_gigi_era

@KhalilBadru

@slowreadersclub

@richvillodas

@sozita_LST

@EastCatholicAD

@tpowersofficial

@funnycrown6

@Courier91711

@telecomparis

@ronromano

@aboalkyan

@secrethazem

@roccoseleven

@secretlab_JP

@syatikunogobou

@JournalStarNews

@BombasGens

@BallasTV

@HazenPolice

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

We actually won the 3DV Best Paper Award for the Grasping Field Paper!!! Super proud of my first year PhD student Korrawe Karunratanakul! Paper:

9

33

378

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

I'm excited to share that our new generative human motion model GAMMA ( #CVPR2022 , with @cnsdqzyz ) has gained life in the exhibition "Motion. Autos, Art, Architecture” at the Guggenheim Museum Bilbao curated by Norman Foster Foundation! (1/3)

3

34

333

Siyu Tang @VLG-ETHZ

@SiyuTang3

11 months

Flight canceled. First time unable to attend CVPR due to visa issues since 2012. Being part of the organization team, I've seen the effort made by the organizers to assist with visa processes. The frustration is shared by everyone involved. IRCC is not doing any favor for Canada.

#CVPR2024

@CVPR

11 months

For several months, the organizers have actively raised concerns with Canadian immigration authorities (IRCC), government agencies, and politicians. In some cases, we have been successful in helping people obtain visas, but in many cases, our efforts were unsuccessful. (2/4)

2

4

28

29

27

307

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

5 papers accepted at #CVPR2021 . Congrats to my students and all the collaborators. Topics are 1. Learning neural articulated occupancy of people; 2. Piecewise transformation field for 3D human body fitting; 3. Marker-based human motion synthesis;

7

3

265

Siyu Tang @VLG-ETHZ

@SiyuTang3

4 months

I’m very happy to share another recent work: Diffusion Noise Optimization (DNO). We demonstrate how to optimize diffusion latent noise using criterion functions defined in motion space, which serves as universal motion priors for a wide range of motion-related tasks.

1

26

227

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

Super excited to share MetaAvatar: Learning Animatable Clothed Human Models from Few Depth Images! We build human avatars from monocular depths or a single scan efficiently, using meta-learned generalizable and controllable neural SDFs. #NeurIPS2021 Code:

0

20

153

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

SCALE models 3D clothed humans with hundreds of articulated surface elements, resulting in avatars with realistic clothing that deforms naturally even in the presence of topological change! #cvpr2021 Awesome work from @qianli_m !

AK

@_akhaliq

3 years

SCALE: Modeling Clothed Humans with a Surface Codec of Articulated Local Elements pdf: abs: project page:

0

11

48

1

20

130

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

We are excited to announce the EgoBody challenge at #ECCV2012 ! The EgoBody benchmark provides pseudo-ground-truth body meshes for natural human-human interaction sequences captured in the egocentric view. Details about the dataset and the challenge:

4

26

119

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

#3DV2021 Oral session 1. Excited to share HALO: a neural implicit representation of human hands that is fully driven by 3D keypoints. A new hand model tailored for interaction modelling. Joint work with Korrawe, @zc_alexfan and @ait_eth Code is here: .

0

12

96

Siyu Tang @VLG-ETHZ

@SiyuTang3

1 year

We are excited to announce the upcoming 3DV 2024 will take place in Davos, Switzerland. Submission: July 31, 2023 Conference: March 18-21, 2024 Location: Davos Congress Centre (the same venue as the annual World Economic Forum) Mark your calendars and join us in Davos! ⛷️🏂

International Conference on 3D Vision

@3DVconf

1 year

📣 Exciting news! 📣 The call for papers is now open for 3DV2024, taking place in the stunning Davos, Switzerland! 📝 Submission: by July 31. 🔗 Details: 🌟 Don't miss the chance to share your research at #3DV2024 🎉 #CallForPapers #Davos #Switzerland

1

37

123

1

9

96

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

Thanks for sharing! Our new #iccv oral paper, LEMO, leverages large-scale Mocap datasets to learn powerful motion priors, and enables accurate human motion capture in 3D scenes with a monocular Kinect. Great collaboration between #Microsoft and @CSatETH !

2

11

89

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

#ECCV2022 . Inhabiting the Virtual. We made another step toward synthesizing virtual humans interacting with 3D scenes. The key idea is to learn a joint representation that effectively captures human body articulation, 3D object geometry, and interaction semantics. (1/2)

AK

@_akhaliq

2 years

Compositional Human-Scene Interaction Synthesis with Semantic Control abs: project page: github:

0

11

60

3

10

85

Siyu Tang @VLG-ETHZ

@SiyuTang3

1 year

🥳SDF Studio is online now: a common framework and an open-source repo for implicit surface reconstruction!

Autonomous Vision Group

@AutoVisionGroup

1 year

🎅 Today, we have an early Christmas present for you: SDF Studio. Building on the fantastic @nerfstudioteam code, we have integrated various implicit surface reconstruction techniques in one common framework! More algorithms and results coming soon..

6

156

593

0

11

85

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

Another #3DV2021 paper: 4D Human Body Capture from Egocentric Video via 3D Scene Grounding. Capturing humans from egocentric videos is a key building block for the embodied and immersive future. Code: . Joint work with Miao, @cnsdqzyz and @RehgJim .

0

14

82

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

I will be looking for PhD students working on neural body modeling for 3D human reconstruction and synthesis. If you are interested in the topic, please consider to apply for the PhD fellowship @ETH_AI_Center @CSatETH

Andreas Krause

@arkrause

3 years

Calls out for Ph.D. and postdoc fellowships in #AI at the @ETH_AI_Center at @ETH_en ! Applications close November 30, 2021. Please spread the word! More information at

3

75

259

1

10

77

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

Neural implicit human body representations provide an efficient way to resolve collisions with objects and 3D scenes. For us, COAP ( #CVPR 2022) is an important step forward, an essential tool to capture and synthesize humans moving in and interacting with complex 3D scenes.

Marko Mihajlovic

@marko_mih

2 years

📢 Our generalizable neural implicit body leverages a localized encoder-decoder to model volumetric humans #CVPR2022 COAP is useful for resolving self-intersections and collisions with other objects w/ @psyth91 @aayushbansal @MZollhoefer @SiyuTang3

0

13

109

1

16

74

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

Great opportunities for exciting AI research! I will be looking for Ph.D. students who are interested in the intersection of Computer Vision, Digital Humans, and Egocentric perception.

ETH AI Center

@ETH_AI_Center

2 years

Apply now and start your PhD or Post-Doc at the ETH AI Center to shape the future of #AI and #ML with interdisciplinary AI research and open your career path to academia, industry and start-ups! Apply before Nov 30: @CSatETH @ETH

3

43

67

2

11

64

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

@cnsdqzyz Successfully inhabiting this six-hundred-meter-high vertical city is a paradigm shift in creating autonomous virtual humans. Flight Assembled City: Human motion model (GAMMA @CVPR22 ): Human body model: SMPL-X from @PerceivingSys

1

11

66

Siyu Tang @VLG-ETHZ

@SiyuTang3

1 month

3DV keynote @BenMildenhall , the live stream is here:

0

5

59

Siyu Tang @VLG-ETHZ

@SiyuTang3

5 months

Very excited to share this work. We achieved a high-quality reconstruction of clothed human avatars with disentangled geometry, albedo, material, and environmental lighting from only a monocular video. Excellent and solid work by @sfwang0928 and the team! #IntrinsicAvatar

Autonomous Vision Group

@AutoVisionGroup

5 months

Given a monocular video, IntrinsicAvatar learns animatable clothed human avatars with decomposed intrinsic properties including albedo, material, and geometry. It was a great @ELLISforEurope collaboration with @SiyuTang3 , @sfwang0928 and @anticboz !

1

13

74

1

55

Siyu Tang @VLG-ETHZ

@SiyuTang3

1 month

Couldn’t agree more!

Andrea Tagliasacchi 🇨🇦🏔️

@taiyasaki

1 month

The @3DVconf PC. Couldn’t have asked for a better team, really 🥰

1

0

75

0

56

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

PTF provides a robust and fully automatic way to fit parametric human models to sparse point cloud, dense scans and monocular depth frames! Paper id: 2691. Q&A Session: 12-14pm (CET) on June 23. #CVPR2021 . with Shaofei from @CSatETH and Andreas @AutoVisionGroup

1

6

55

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

The power of points!

AK

@_akhaliq

3 years

The Power of Points for Modeling Humans in Clothing pdf: abs: project page:

0

11

50

2

7

55

Siyu Tang @VLG-ETHZ

@SiyuTang3

1 year

Excited to share Mask3D - a new state-of-the-art for 3D instance segmentation on Point Clouds on ScanNet. #ICRA2023 Online demo: . You can test your own point clouds and download the results!

Francis Engelmann

@FrancisEngelman

1 year

Mask3D 🎭 is now at #ICRA2023 , great work @JonasSchultCV ! We use Mask Transformers for 3D Instance Segmentation on Point Clouds ~ 🥇 on ScanNet 📰Paper: 🛠️Project: 👨‍💻Code: @Pandoro89 @orlitany @SiyuTang3

3

51

205

0

8

52

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

Check out our new #3dv2020 oral paper: Grasping Field: Learning Implicit Representations for Human Grasps. Time (CET): Friday: 6 - 6.30 AM 16.30-17.00 PM Paper: Video: code:

GitHub - korrawe/grasping_field

Contribute to korrawe/grasping_field development by creating an account on GitHub.

github.com

0

4

52

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

It was super fun! Glad to be there! I talked about our recent work on capturing and modeling of 3D humans. Here is the slides (including an overview of our research):

Digital Festival 2021(Zürich)

The presentation at the Digital Festival 2021 (no audio)

www.youtube.com

Joël Mesot

@Joel_Mesot

3 years

Yesterday the @ETH_en session on augmented reality glasses with @SiyuTang3 & @ait_eth was well attended at the Digital Festival @DiFe_Zurich #DiFe21 , after a fascinating address by Professor Ash @ellliottt on #AIsystems earlier in the day #AR #AI #AIglasses

1

3

26

0

3

52

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

Going to give a talk on neural bodies and hands for interaction capture and synthesis at the ICCV workshop SoMoF, hope to see you there: 😀 Papers will be covered:

1

5

47

Siyu Tang @VLG-ETHZ

@SiyuTang3

11 months

Excited to attend and present at ICRA for the very first time! Dora’s paper, Interactive Object Segmentation in 3D Point Clouds, is one of the outstanding paper finalists for physical human-robot interaction at ICRA 2023! 🥳

TheodoraKontogianni

@DoraKontog

11 months

Happy to be at #ICRA2023 !Please come and see our work ( @SiyuTang3 ,Konrad Schindler, @ekincelli ) during the poster session (30th May, Pod 12, 15:00-16:40) and the oral presentation (31st May, ICC Auditorium, 10:20-10:30) 📜 👩‍💻

0

3

34

0

4

43

Siyu Tang @VLG-ETHZ

@SiyuTang3

1 year

Want to give a comprehensive overview of a specific topic that is related to CV at @CVPR 2023? The deadline for submitting a tutorial proposal is fast approaching! (Dec 9, 2022) Details: For any questions, please contact me and JianxinWu ( tutorial chairs)

0

6

41

Siyu Tang @VLG-ETHZ

@SiyuTang3

1 year

It was a great honor to present our research from VLG at BMVC 2022. Thanks for the invitation!

Constantino Carlos Reyes-Aldasoro

@C_ReyesAldasoro

1 year

Very engaging keynote presentation on virtual humans by Prof Siyu Tang @SiyuTang3 #bmvc2022 @TheBMVA

1

0

6

1

0

41

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

LEAP (4356) provides neural occupancy representations for SMPL bodies. Paper Session 8 (Wed, 10 PM EDT / Thur, 4 AM CET). Project & code: with @marko_mih @cnsdqzyz from @CSatETH and @Michael_J_Black !

2

5

41

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

If you are attending #3dv2020 , please stop by our poster: PLACE: Proximity Learning of Articulation and Contact in 3D Environments, with @Michael_J_Black @qianli_m Yan zhang and Siwei Time: 5.30-7.00 pm (CET) Code:

0

6

39

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

This video summarizes the #ECCV2022 papers we are going to present in the next three days. If you are interested in egocentric human body estimation, scalable human avatar creation, and human motion and interaction synthesis, don't miss it. @eccvconf

ECCV 2022 Computer Vision and Learning Group (VLG) at ETH Zurich

null

www.youtube.com

1

5

33

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

I will give a spotlight presentation about the latest work from my group, including neural body representation, human motion modelling, human body estimation from sparse point cloud, and human grasps synthesis. Hope to see you at the event!

ETH CS Department

@CSatETH

3 years

🎉 6 Months ETH AI Center! April 15, 2021, 17:00 (ECT): Celebration of Six Months ETH AI Center and AI+X Kick-off, registration + link to livestream here @ETH_en @ryandcotterell @SiyuTang3 @SwissCognitive #AI #Foresight

0

8

20

0

33

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

In MOJO, we propose a novel representation of the body in motion, a novel motion generative network, and a novel scheme for 3D body mesh recovery. #cvpr2021 Great work from @cnsdqzyz !

Michael Black

@Michael_J_Black

3 years

MOJO predicts 3D human movement. Previous work predicts 3D joints, which are just sparse point clouds. Instead, we predict a point cloud of body surface points and ensure that they correspond to a true 3D body. @CVPR ( #CVPR2021 ).

0

8

61

2

4

32

Siyu Tang @VLG-ETHZ

@SiyuTang3

1 year

Thank you very much for the invitation!! The view from Cabot Tower was fantastic🥳🥳 I learned so much about egocentric vision from your group!😃

Dima Damen

@dimadamen

1 year

"Inhibiting the Virtual"... @SiyuTang3 thanks for your visit to give a fantastic talk at MaVi seminar @BristolUniEng @bristolcs and spending the day hearing about our research. The weather didn't disappoint - typical gloomy but we managed to get up #cabotTower #Bristol

0

1

23

0

30

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

GAMMA ( #CVPR2022 ) is now deployed for HoloLens. Autonomous virtual humans are coming for AR applications.

[CVPR 2022] The Wanderings of Odysseus in 3D Scenes

GAMMA is a solution to animating 3D bodies.

yz-cnsdqz.github.io

Yan Zhang

@cnsdqzyz

2 years

Can you imagine how virtual humans wander in the ETH main building? Based on SMPL-X, Hololens, and our method GAMMA(), now we have it! Stay tuned...

1

4

30

0

4

28

Siyu Tang @VLG-ETHZ

@SiyuTang3

10 months

Congratulations to @SiweiZhang13 🥳!! Well deserved! #VLG is very proud of you!

Qualcomm Research & Technologies

@QCOMResearch

10 months

QIF Europe is an excellence award through which @Qualcomm rewards and mentors the most innovative PhD students in Europe working on breakthrough #AI and #cybersecurity solutions. Congratulations @tychovdo @confusezius @SiweiZhang13 and Attri Bhattacharyya

0

7

26

3

28

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

LEAP: Learning Articulated Occupancy of People #CVPR2021 . It is a neural network architecture for representing volumetric animatable human bodies. Project page: Great work from @marko_mih and congrats to the team!! @cnsdqzyz @Michael_J_Black

[CVPR 2021] LEAP: Learning Articulated Occupancy of People

LEAP is a neural network architecture for representing volumetric animatable human bodies.

neuralbodies.github.io

1

28

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

Capturing human motion and activities from egocentric videos is a key building block for our embodied and immersive future. Come to join us tomorrow at 2 pm. The HBHA workshop features three benchmarks (EgoBody, H2O, and Assembly101) and has 5 amazing speakers. @eccvconf

Siwei Zhang

@SiweiZhang13

2 years

Our HBHA workshop on human body, hands and activities from egocentric and multi-view cameras will take place at #ECCV2022 on Monday (Oct.24) 2pm-6pm (GMT+3)! Don’t miss out the invited talks from 5 wonderful speakers @dimadamen @gulvarol @RehgJim @kkitani and Vincent Lepetit!

2

8

47

0

2

26

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

I'm very excited to work on this project with @CSatETH , @Microsoft , @derbalgrist , and the team!

Marc Pollefeys

@mapo1

2 years

Excited to participate in this project with the Microsoft Mixed Reality and AI Zurich Lab and collaborate closely with Philipp Fürnstahl, @SiyuTang3 , Grabner Helmut and teams. #microsoft #ai #mixedreality #mesh #azure #ethzurich #balgrist #zhaw

0

4

28

0

1

27

Siyu Tang @VLG-ETHZ

@SiyuTang3

1 year

Great opportunity! Andreas is a brilliant supervisor, I cannot recommend him highly enough.

Autonomous Vision Group

@AutoVisionGroup

1 year

I am hiring PhD students and PostDocs! If you like to join a great team to conduct curiosity-driven research on implicit neural 3D representations join us now! Flyer:

11

74

378

0

3

26

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

Our human motion priors for 4D human motion and interaction capture in 3D scenes will be presented today. #ICCV2021 oral. Session 9. Code is available: Video:

LEMO: Learning Motion Priors for 4D Human Body Capture in 3D Scenes

Recovering high-quality 3D human motion in complex scenes from monocular videos is important for many appli- cations, ranging from AR/VR to robotics. However...

www.youtube.com

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

Thanks for sharing! Our new #iccv oral paper, LEMO, leverages large-scale Mocap datasets to learn powerful motion priors, and enables accurate human motion capture in 3D scenes with a monocular Kinect. Great collaboration between #Microsoft and @CSatETH !

2

11

89

0

25

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

Happy to share PTFs: Locally Aware Piecewise Transformation Fields for 3D Human Mesh Registration. #cvpr2021 . PTFs simultaneously learn to predict shape and per-point correspondences for sparse point clouds of humans. Video:

[CVPR 2021] Locally Aware Piecewise Transformation Fields for 3D...

Registering point clouds of dressed humans to paramet- ric human models is a challenging task in computer vi- sion. Traditional approaches often rely on heav...

www.youtube.com

1

5

25

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

Self-contact is ubiquitous in human behavior, in TUCH, we introduce the first human pose and shape regressor for self-contact poses, along with a novel dataset of 3D human meshes with realistic contact. #cvpr2021 oral. Great work from @LeaMue27 !!!

Michael Black

@Michael_J_Black

3 years

TUCH: training human pose and shape regression with novel *self-contact* losses improves accuracy even on poses without self-contact. Mimic-The-Pose (MTP): novel crowd-sourced, high-quality, 3D reference data with self-contact. @CVPR ( #CVPR2021 oral).

3

13

95

1

4

24

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

KeypointNerf ( #ECCV2022 ) enables novel view synthesis for human faces and bodies using only keypoints, without mesh-based parametric models. Marko talked about this work in great detail in this podcast.

Talking Papers Podcast

@talking_papers

2 years

@marko_mih is a 2nd year PhD student at ETH, supervised by @SiyuTang3 . His research focuses on photorealistic reconstruction of static and dynamic scenes and also modeling of parametric human bodies. This work was done mainly during his internship at Meta Reality Labs.

1

0

4

0

1

23

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

I will present this work today at #CVPR2022 Session 4.2 Poster 172 14:30AM-17:00PM, happy to chat in person during the session for more details! Code is available: .

[CVPR 2022] The Wanderings of Odysseus in 3D Scenes

GAMMA is a solution to animating 3D bodies.

yz-cnsdqz.github.io

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

I'm excited to share that our new generative human motion model GAMMA ( #CVPR2022 , with @cnsdqzyz ) has gained life in the exhibition "Motion. Autos, Art, Architecture” at the Guggenheim Museum Bilbao curated by Norman Foster Foundation! (1/3)

3

34

333

0

3

23

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

Wonderful news! Congrats!!!

ICCV2021

@ICCV_2021

3 years

Best Student Paper @ICCV_2021 Pixel-Perfect Structure-from-Motion with Featuremetric Refinement Philipp Lindenberger (ETH Zurich), Paul-Edouard Sarlin (ETH Zurich), Viktor Larsson, Marc Pollefeys [Session 5 A/B]

1

10

60

0

20

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

If you missed it, here is the recording of the VLG part (Capture and Synthesis of 3D Humans):

VLG spotlight at the Celebration of Six Months ETH AI Center

I gave a spotlight presentation at the celebration of six months ETH AI Center. I talked about the latest works from VLG.

www.youtube.com

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

I will give a spotlight presentation about the latest work from my group, including neural body representation, human motion modelling, human body estimation from sparse point cloud, and human grasps synthesis. Hope to see you at the event!

0

33

0

1

20

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

Joint work with Shaofei, @marko_mih @qianli_m and @AutoVisionGroup 🥳🥳🥳

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

Super excited to share MetaAvatar: Learning Animatable Clothed Human Models from Few Depth Images! We build human avatars from monocular depths or a single scan efficiently, using meta-learned generalizable and controllable neural SDFs. #NeurIPS2021 Code:

0

20

153

0

1

19

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

Congratulations!!!

Autonomous Vision Group

@AutoVisionGroup

3 years

This is unbelievable: Michael Niemeyer @Mi_Niemeyer won the CVPR 2021 best paper award!! Thanks to Michael and all my fantastic students and postdocs for their extraordinary work! You are just amazing..

45

49

577

0

2

19

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

Thrilled that our interdisciplinary collaboration with Gramazio Kohler Research on inhabiting a vertical city with autonomous virtual humans is featured on the ETH front page! Joint work with @cnsdqzyz and Jonathan

ETH Zurich

@ETH_en

2 years

What happened if architects let their buildings be tested by virtual inhabitants before actually building them? A museum project by Matthias Kohler, @SiyuTang3 and Fabio Gramazio could serve as a blueprint: @CSatETH #Architecture #ComputerVision

0

4

14

0

1

16

Siyu Tang @VLG-ETHZ

@SiyuTang3

11 months

A fantastic list of speakers covering super exciting topics in the incredible city of Paris. Don't miss out 🥳

Despoina Paschalidou

@paschalidoud_1

11 months

📢Our #ICCV2023 workshop on AI for 3D Content Creation organized with @geopavlakos , @amlankar95 , @KaichunMo and @davrempe from, Paul Guerrero, @SiyuTang3 and Leo Guibas has a fantastic list of speakers! Workshop Website: Paper Submission Deadline: July 17

3

24

147

0

2

16

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

MOJO is a surface marker based human motion prediction model. Our recursive projection scheme supported by the marker-based representation directly yields body model parameters and hence realistic body meshes. The paper [2253] will be presented today at 12-14:30pm (CET)

Yan Zhang

@cnsdqzyz

3 years

#CVPR21 #CVPR2021 #CVPR please come to our paper MOJO [2253] with @Michael_J_Black and @SiyuTang3 at (CET) 12:00– 14:30 on June 22, or (EDT) 6:00– 8:30 on June 22. 👇CODE IS RELEASED. Please go to or scan it for everything.👇

2

3

25

0

1

15

Siyu Tang @VLG-ETHZ

@SiyuTang3

11 months

What a wonderful retreat! 🥳 Absolutely loved the hike. Thanks so much for the invitation!

Imagine-ENPC

@ImagineEnpc

11 months

Lucky to have two great invited talks from @chriswolfvision and @SiyuTang3 !

2

1

15

0

15

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

I think the articulated local elements are powerful representations for building clothed human avatars. Paper session: 12-14pm (CET) on June 25. #CVPR2021

Qianli Ma

@qianli_m

3 years

@cnsdqzyz Thanks! A point is not guaranteed to always be the same physical (semantic) point on the surface across different frames, as we allow the patches to freely move. But as the network also predicts point colors (the one in the middle), the results still look coherent across frames.

0

1

0

14

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

By leveraging the attention mechanism and transformer, our method can synthesize compositional human-scene interactions without requiring composite interaction data. (2/2) Joint work with Kaifeng, @sfwang0928 @cnsdqzyz , and Thabo. @CSatETH Code:

GitHub - zkf1997/COINS: code release of paper "Compositional Human-Scene Interaction Synthesis with...

code release of paper "Compositional Human-Scene Interaction Synthesis with Semantic Control" - zkf1997/COINS

github.com

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

#ECCV2022 . Inhabiting the Virtual. We made another step toward synthesizing virtual humans interacting with 3D scenes. The key idea is to learn a joint representation that effectively captures human body articulation, 3D object geometry, and interaction semantics. (1/2)

3

10

85

0

1

14

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

KeypointICON 🥳

Yuliang Xiu

@yuliangxiu

2 years

KeypointNeRF's "relative spatial keypoint encoder" is a general plug-n-play module for different downstream tasks. I have integrated it with ICON, which achieves comparable performance, compared with expensive body SDF. More details at:

1

2

22

0

13

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

Very happy to hear your comments! It was super fun and we had an exciting panel discussion with @SattlerTorsten @ait_eth @GerardPonsMoll1 @angelaqdai

Michael Black

@Michael_J_Black

3 years

Folks, this was a great talk -- inspiring and clear. If you missed it, you can still watch it on YouTube

1

3

31

0

1

12

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

This is a great opportunity! Dimitris is a great person to work with!

Dimitris Tzionas

@dimtzionas

2 years

Motivated MSc/BSc students & prospective PhD candidates can always reach out to me -- plz see the contact instructions on my website. Plz help me spread the word: 🆘 🆘 🆘 We are actively hiring 1 PhD candidate together with @theogevers ! Separate tweet coming soon 📢 📢 📢 (8/8)

5

18

0

12

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

@cnsdqzyz Together with Gramazio Kohler Research, we developed a visionary project in which an entirely pedestrianized city is populated with many autonomous virtual humans who possess diverse body shapes and move perpetually in an automatic, scalable, and controllable manner. (2/3)

1

2

11

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

4, Clothed human with Surface Codec of Articulated Local Elements; and 5 on self contact and human pose!

1

0

11

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

Looking forward to this #CVPR2022 tutorial! A great opportunity to learn more about biological vision.

Li Zhaoping

@Li_Zhaoping

2 years

@CVPR #CVPR2022 I am giving a CVPR2022 tutorial "A post-Marrian computational overview of how biological (human) vision works", see could adjust (a bit) some contents according to participants' interest, let me know.

0

1

8

1

0

10

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

🥳🥳🥳

Michael Black

@Michael_J_Black

3 years

Neural people are coming! SCANimate, SCALE, and now LEAP - all at #CVPR2021 . We are exploring how to learn 3D humans that go beyond models like SMPL. Great collaboration between ETH and MPI with @marko_mih , @cnsdqzyz and @SiyuTang3 . (1/4)

1

7

49

0

9

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

I'm looking forward to this colloquium. It provides us an opportunity to discuss how our human motion and interaction models could improve the sustainability of the built environment with the AEC researchers. Virtual humans may help us to build a more sustainable real world.

Iro Armeni

@ir0armeni

3 years

It will be held as a series of online sessions: the 1st one will focus on building reuse & will happen on November 23rd 4-7 PM CET. We are excited to have @billmcdonough give a keynote talk, as well as Ruchi Choudhary, Clara Olóriz Sanjuán, & @SiyuTang3 present their perspective.

0

1

4

0

1

8

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

Congrats to the team @SiweiZhang13 , @cnsdqzyz , and the amazing collaborators @FedericaBogo , @mapo1

0

7

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

Looking forward to this workshop. @SiweiZhang13 will also present our EgoBody dataset there!

Ego4D

@ego4_d

2 years

Challenges! We concluded 16 different Ego4D challenges earlier this month and @_rohitgirdhar_ will share for the first time the winners and provide a synthesis on what methods were effective. 🤓🧐

1

0

2

0

2

7

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

Great collaboration with @psyth91 @aayushbansal @MZollhoefer . Proud of @marko_mih from @CSatETH .

0

7

Siyu Tang @VLG-ETHZ

@SiyuTang3

4 months

It is so simple that it can be used directly without per-task training or finetuning, and in this paper, we show how to use it for motion editing, motion in-between, motion denoising, and motion blending with exactly the same algorithm!

1

0

5

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

The egocentric human capture and reconstruction weren't easy! Great work by @SiweiZhang13 and the collaborators @mapo1 @FedericaBogo @qianli_m @cnsdqzyz @big_stamp @CSatETH @ETH_AI_Center .

0

5

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

Great collaboration between VLG and AVG 🥳🥳🥳

Autonomous Vision Group

@AutoVisionGroup

3 years

In our upcoming CVPR'21 paper (joint work with Shaofei and @SiyuTang3 ), we propose Piecewise Transformation Fields (PTF) to simultaneously learn to predict human shape and per-point correspondences.

0

3

23

0

4

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 months

@chriswolfvision @3DVconf @iclr_conf This wasn't intended😂

0

4

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

@anaklimovic @Facebook Congrats Ana!!!

0

4

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

More information and baseline results can be found in our ECCV paper: If you are interested in egocentric human motion capture and understanding, download the data and give it a try — the submission deadline is October 1st.

2

0

4

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 months

@taiyasaki @CSProfKGD Absolutely :)

1

0

2

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

@KhrylxYe @jessicahodgins @davheld @FidlerSanja @kkitani Congrats!!

1

0

2

Siyu Tang @VLG-ETHZ

@SiyuTang3

4 months

Project page: Paper: Excellent work by Korrawe, and the team @phizaz @emreaksan , Thabo, and Supasorn

0

1

2

Siyu Tang @VLG-ETHZ

@SiyuTang3

1 year

@taiyasaki @jbhuang0604 @majti89 @mapo1

0

2

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

@geopavlakos @akanazawa super cool work!

0

2

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

@JamieDJS @Awfidius @tobysharp @akipman Congratulations!!!

0

2

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

code is released on Github:

0

2

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

There is a typo. The EgoBody paper and challenge will be presented at #ECCV2022

0

2

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

Consequently, rigid per bone transformations and joint rotations can be obtained efficiently via a least-square fitting given the estimated point correspondences, circumventing the challenging task of directly regressing joint rotations from neural networks.

0

2

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

@Michael_J_Black thanks!! Michael!

0

1

Siyu Tang @VLG-ETHZ

@SiyuTang3

2 years

@mohomran Congratulations!!! 🥳🥳🥳

1

0

1

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

project page:

1

0

1

Siyu Tang @VLG-ETHZ

@SiyuTang3

1 year

@M_E_Hassan Congratulations!!!

0

1

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

@Michael_J_Black @Leopoldina Congratulations!

0

1

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

Our key insight is that the translation vector for each query point can be effectively estimated using the point-aligned local features.

1

0

1

Siyu Tang @VLG-ETHZ

@SiyuTang3

3 years

@GerardPonsMoll1 looking forward to reading these papers!

0

1