JavaScript is not available.

Alexia Jolicoeur-Martineau

@jm_alexia

Dec 23, 2022

I just discovered 'torchview', a cool new python library to visualize the graph of a neural network in Torch. It seems a lot better than previous existing libraries, it shows you all the important details without the fluff.

github.com

GitHub - mert-kurttutan/torchview: torchview: visualize pytorch models

torchview: visualize pytorch models. Contribute to mert-kurttutan/torchview development by creating an account on GitHub.

5

67

474

Dec 20, 2022

This is great news! Hope to see similar announcements for graduate students in

Quote Tweet

Temerty Medicine is substantially increasing its graduate student stipends beginning in the 2023-2024 academic year, making them the highest in Canada. bit.ly/3HQRFex

4

.

Quote Tweet

Caltech sets minimum 23-24 grad school stipends at $45K.

2

11

José M. Álvarez on LinkedIn: #nvidia #intern #neurips #icml #iclr #cvpr #eccv #iccv #nerf #pruning…...

Jose M. Alvarez

@ALVAREZ_JOSEM

Dec 8, 2022

Our Autonomous Vehicle Perception Research team at #nvidia has several year-round openings for strong research #intern (PhD and MS, preferably). Ideally, a candidate with proven track record: research publications in top-tier venues such as #neurips , #i…

linkedin.com

Our Autonomous Vehicle Perception Research team at #nvidia has several year-round openings for strong research #intern (PhD and MS, preferably). Ideally, a… | 18 comments on LinkedIn

6

19

Math for Deep Learning — Andreas Geiger

Autonomous Vision Group

@AutoVisionGroup

Dec 7, 2022

To level student's background knowledge for our deep learning lecture, I recorded a series of micro tutorials on relevant concepts (linear algebra, differential calculus, probability/information theory). Slides: drive.google.com/file/d/1SecJqk Videos:

youtube.com

Lecture: Math for Deep Learning (MaDL) A Short Guided Tour on Linear Algebra and Probability Theory (Prof. Andreas Geiger, University of Tübingen) Course Web...

3

69

369

PyTorch

@PyTorch

Dec 2, 2022

We just introduced PyTorch 2.0 at the #PyTorchConference, introducing torch.compile! Available in the nightlies today, stable release Early March 2023. Read the full post: bit.ly/3VNysOA 🧵below! 1/5

22

626

1,911

Adobe Research

@AdobeResearch

Dec 1, 2022

Adobe researchers are presenting new work at this year's Neural Information Processing Systems conference (#NeurIPS2022). Check out the list of accepted papers and other contributions from

@Adobe

! adobe.ly/3itKoXA

2

12

31

Guandao Yang

@stevenygd

Tired of using neural fields as black boxes? Want provable guarantees in the output of your neural fields? Check out our NeurIPS paper: Polynomial Neural Fields! Poster time: Tue 29 Nov 2 p.m. PST - 4 p.m. PST Poster location: Hall J #904 Paper link: openreview.net/pdf?id=juE5Erm

GIF

24

94

Peyman 𝕄𝕀𝕃𝔸ℕ𝔽𝔸ℝ

@docmilanfar

Want to work @ the frontiers of Computational Photography & Generative Models? Our team is hiring interns. DM me or

@2ptmvd

& apply now: 1. PhD Research Interns (final yr students) careers.google.com/jobs/results/1 2. Student Researchers (graduating by Dec 2024) careers.google.com/jobs/results/1

3

21

101

Vector Institute

@VectorInst

Have you heard about Breaking Bad? Well, this isn't the TV show, this type includes a large-scale dataset of fractured objects. Read

@sellan_s

, Yun-Chun Chen, Ziyi Wu,

@animesh_garg

+

@_AlecJacobson

highlighted paper below👇 #VectorAtNeurIPS #NeurIPS2022 openreview.net/pdf?id=mJWt6pO

1

7

15

We will be presenting our

@NeurIPSConf

paper today from 11 am to 1 pm CST at Hall J #1024. Come talk to us to learn about the Breaking Bad dataset. breaking-bad-dataset.github.io

Quote Tweet

I’m excited to share our @NeurIPSConf paper: Breaking Bad: A Dataset for Geometric Fracture and Reassembly. This is joint work with @sellan_s, @Dazitu_616, @animesh_garg, and @_AlecJacobson. Paper: arxiv.org/abs/2210.11463 Project page: breaking-bad-dataset.github.io (1/12)

Show this thread

2

3

Kosta Derpanis

@CSProfKGD

Nov 28, 2022

Here's something you don't see often, a faculty-only authored paper.

@angelaqdai

@MattNiessner

arxiv.org/abs/2211.14249

3

27

200

The Artificial Intelligence (AI) Residency Program - Meta AI

Roozbeh Mottaghi

@RoozbehMottaghi

Nov 28, 2022

My team at FAIR has multiple positions for AI residents in 2023. Our focus is on Embodied AI and Perception. If you are interested, please email me. Here is the application link:

ai.facebook.com

The Artificial Intelligence (AI) Residency Program is a one-year research training position designed to give you hands-on experience with artificial intelligence research while working in Meta AI.

2

22

150

Christian Wolf

@chriswolfvision

Nov 24, 2022

TV shows inspired papers in ML+CV: R2D2: arxiv.org/abs/1906.06195 LOFTR: arxiv.org/abs/2104.00680 Breaking bad: arxiv.org/abs/2210.11463 Game of Thrones: arxiv.org/abs/1810.11162 Star Trek: arxiv.org/abs/2108.09904

9

11

44

Joey Burant

@jbburant

Nov 23, 2022

#GradStudents are arguably *the* primary producers of #research output at Canadian universities, and core contributors to the #teaching and #leadership missions of their institutions. **But how (and how well) are they compensated?** Let's take a look! A long thread 🧵👇 (1/n)

12

208

492

Alec Jacobson

@_AlecJacobson

Nov 20, 2022

Conformal and globally bijective parametrization for sphere-topology surfaces with a single sparse linear system solve! The cuts connecting selected cone singularities are irrelevant because the mapping seamlessly tiles the plane. Orbifold Tutte Embeddings by Aigerman & Lipman

GIF

2

22

123

Peter Jenks

@psejenks

Nov 14, 2022

Starting today I'm cancelling live instruction in solidarity with the striking

@UCBerkeley

graduate students, postdocs, and researchers. TA/Grad pay is an equity and justice issue: until grad students are paid more, grad school will remain an intolerably risky bet for too many.

20

279

2,606

Andrea Tagliasacchi @ Vancouver

@taiyasaki

Nov 9, 2022

📢📢📢 Excited to announce “𝐧𝐞𝐫𝐟𝟐𝐧𝐞𝐫𝐟: 𝐏𝐚𝐢𝐫𝐰𝐢𝐬𝐞 𝐑𝐞𝐠𝐢𝐬𝐭𝐫𝐚𝐭𝐢𝐨𝐧 𝐨𝐟 𝐍𝐞𝐮𝐫𝐚𝐥 𝐑𝐚𝐝𝐢𝐚𝐧𝐜𝐞 𝐅𝐢𝐞𝐥𝐝𝐬” → PDF: arxiv.org/abs/2211.01600 → Homepage: nerf2nerf.github.io → Code: github.com/nerf2nerf/nerf → Video: youtu.be/S071rGezdNM

0:05

9.8K views

2

55

233

Quote Tweet

I'm excited to share our #RSS2022 workshop paper: Neural Motion Fields: Encoding Grasp Trajectories as Implicit Value Functions. w/ @Adithya_Murali_ , @balakumar_ , Wei Yang, @animesh_garg, and Dieter Fox at @nvidia. Paper: arxiv.org/abs/2206.14854 Video: youtu.be/B-pEhT1pi-Q

Show this thread

1

A very interesting paper! Wonder how this paper compares with the Neural Motion Fields paper (arxiv.org/abs/2206.14854) by

@Adithya_Murali_

,

@balakumar_

,

@animesh_garg

, Dieter Fox, and myself.

Quote Tweet

Thomas Weng

@thomas_weng

Nov 9, 2022

Introducing Neural Grasp Distance Fields, which represent 6-DOF grasps as @neural_fields, enabling joint grasp and motion planning. sites.google.com/view/neural-gr Prior works use discrete grasp sets & multi-stage pipelines. w/@davheld, @_kainoa_, @mhmukadam (@MetaAI, @CMU_Robotics)

Show this thread

GIF

1

5

Animesh Garg

@animesh_garg

Geometric Reasoning is at the core of multiple problems in robotics and vision - assembly, archaeology, and medicine It takes Neural Shape Mating to extremes with shattered objects Check out our #Neurips2022 paper "Breaking Bad" 🧵👇

Quote Tweet

I’m excited to share our @NeurIPSConf paper: Breaking Bad: A Dataset for Geometric Fracture and Reassembly. This is joint work with @sellan_s, @Dazitu_616, @animesh_garg, and @_AlecJacobson. Paper: arxiv.org/abs/2210.11463 Project page: breaking-bad-dataset.github.io (1/12)

Show this thread

5

43

Ziyi Wu

@Dazitu_616

An unforgettable collaboration with wonderful people. This is also my first (co-)first author paper published at top conferences. Come and check our dataset!

Quote Tweet

I’m excited to share our @NeurIPSConf paper: Breaking Bad: A Dataset for Geometric Fracture and Reassembly. This is joint work with @sellan_s, @Dazitu_616, @animesh_garg, and @_AlecJacobson. Paper: arxiv.org/abs/2210.11463 Project page: breaking-bad-dataset.github.io (1/12)

Show this thread

3

43

Silvia Sellán

@sellan_s

Oct 31, 2022

This was a really fun collaboration to work on, and a very poetic one: Using our capacity for breaking things to learn how to put them back together. It was also a great pleasure to work with incredibly hard working students like

@Dazitu_616

and

. Off to New Orleans!

Quote Tweet

I’m excited to share our @NeurIPSConf paper: Breaking Bad: A Dataset for Geometric Fracture and Reassembly. This is joint work with @sellan_s, @Dazitu_616, @animesh_garg, and @_AlecJacobson. Paper: arxiv.org/abs/2210.11463 Project page: breaking-bad-dataset.github.io (1/12)

Show this thread

1

15

arXiv Daily

@Arxiv_Daily

Oct 22, 2022

Breaking Bad: A Dataset for Geometric Fracture and Reassembly deepai.org/publication/br by Silvia Sellán et al. including

#DeepLearning #ComputerScience

,

@Dazitu_616

deepai.org

Breaking Bad: A Dataset for Geometric Fracture and Reassembly

10/20/22 - We introduce Breaking Bad, a large-scale dataset of fractured objects. Our dataset consists of over one million fractured objects ...

3

5

Check out our paper and the following links to learn more! Look forward to introducing our paper at

@NeurIPSConf

this November & December. (12/12)

3

* Future Work * There are a number of directions and problem domains for future work that are made possible by our dataset, including robotics, archaeology, computer graphics, and many more! (11/12)

1

* Why do part assembly methods fail? * Semantic assembly methods leverage part priors by learning global features. However, each fracture does not have a well-defined semantic meaning. Assembling fractures has to rely on learning local features for surface matching. (10/12)

1

* How well do baselines perform? * We evaluate 3 baselines on the everyday subset. DGL, a state-of-the-art part assembly method, performs the best both quantitatively and qualitatively. All methods fail drastically in estimating poses and are far from solving the task. (9/12)

1

2

* Baselines * We benchmark 3 state-of-the-art shape assembly deep learning algorithms on Breaking Bad under various settings. Here shows an overview of a typical vision-based shape assembly pipeline. The model takes as input point clouds and predicts respective poses. (8/12)

1

2

* Dataset gallery * Breaking Bad contains 1 million fractured objects, which can be used for various applications, including artifacts for archaeology, everyday objects for vision and robotics, and other objects for video gaming and example-based fracture simulation. (7/12)

1

2

* Shape selection * We collect base models from PartNet and Thingi10K and apply Sellán et al’s fracture simulation algorithm to each. (6/12)

1

* How do we generate fractured objects? * We adopt a physically based algorithm from Sellán et al. that efficiently generates fractures. For details of fracture simulation, see twitter.com/sellan_s/statu (5/12)

GIF

Quote Tweet

Silvia Sellán

@sellan_s

Jul 18, 2022

I am beyond excited to finally share that our newest paper, "Breaking Good: Fracture Modes for Realtime Destruction" will be presented at SIGGRAPH Asia 2022. Video: youtu.be/0k_tEk34nJQ Paper: silviasellan.com/pdf/papers/fra Code: github.com/sgsellan/fract

Show this thread

1

2

* Semantic ≠ geometric * Objects in existing datasets are decomposed in a semantically consistent way. However, objects that break naturally due to external forces generally do not break into fragments that are semantically well-defined. (4/12)

1

2

* Why a new dataset? * Machine learning approaches for assembly require large-scale datasets of fractured objects. Existing assembly datasets are constructed based on human or automated semantic segmentation. (3/12)

1

4

* What is the task? * Fracture reassembly aims to compose the fragments of a fractured object back into its original shape. (2/12)

GIF

2

1

7

Silvia Sellán and 4 others

Papers with Datasets

@paperswithdata

Oct 28, 2022

🏺Breaking Bad: a large-scale dataset of fractured objects. It has 10k meshes, each with 100 fractures mode, totalling 1M breakdown patterns. It serves as a benchmark for fractured object reassembly and new challenges for geometric shape understanding. paperswithcode.com/dataset/breaki

10

43