[2018.01.11] Facebook

[unizard]

I wonder which topic is studied on Facebook. Let's Go.

https://research.fb.com/ https://github.com/facebookresearch

Research Area + Leader

Direction: Toward AI Agent that can see, talk & act...

• FAIR

  • Yann Lecun, Larry Zitnick, Antoine Bordes, Holger Schwenk
  • https://research.fb.com/category/facebook-ai-research-fair/
  • 전체적인 느낌은 1년에 논문을 5편 이상 쓰는 분들이 가는 곳으로 보여짐... ㅋㅋㅋㅋㅋㅋ

• Compuver VIsion

  • Richard Szeliski(RE), Michael Cohen(R) + Manohar Paluri(RE), Jamie Ray(E)
  • https://research.fb.com/category/computer-vision/

• Applied Machine Learning (AML)

  • Joaquin Qunonero Candela, Richard Szeliski, Tommer Leyvand, Reena Philip
  • https://research.fb.com/category/applied-machine-learning/

• VR

  • Ahannon Chen
  • https://research.fb.com/category/virtual-reality/

FAIR-FB Director & Leader

• Yann LeCun, Director, NYU Professor, Menlo Park, NY, Paris

  • ML, DL, CV
  • http://yann.lecun.com/

• Joaquin Qunonero Candela, Director, Menlo Park (Ph.D 04)

  • AML, ML
  • http://quinonero.net/

• Larry Zitnick, Lead, UW Professor, Menlo Park

  • CV
  • http://larryzitnick.org/

• Antoine Bordes, Lead, NY

  • NLP
  • https://www.hds.utc.fr/~bordesan/dokuwiki/doku.php

• Holger Schwenk, Lead, Paris

  • ML, NLP
  • http://www-lium.univ-lemans.fr/~schwenk/

• Richard Szeliski, Director, WA Professor, Menlo Park

  • CV, CG
  • Director of Computational Photography group
  • https://research.fb.com/people/szeliski-richard/

• Michael Cohen, Lead, (FB 15)

  • AML

• Manohar Paluri, Lead, Menlo Park

  • CV, Engineer
  • https://sites.google.com/site/balamanohar/
  • My group focuses on building the image and video understanding platform used by all Facebook and Instagram product groups.

[unizard]

FAIR, Montreal, ML, 9/15/2017

• Joelle Pineau, Montreal (Visiting Researcher), ML

  • http://cs.mcgill.ca/~jpineau/

• Mike Rabbat, Montreal, ML

  • https://research.fb.com/people/rabbat-mike/

• Pascal Vincent, Montreal

  • https://research.fb.com/people/vincent-pascal/

[unizard]

FAIR, Paris, CV 6/2/2015

• Camille Couprie (LImited Anotated Data), France (Ph.D 11) (FB 15)

  • FAIR (CV)
  • https://perso.esiee.fr/~coupriec/

• Armand Joulin, France (Ph.D 12) (FB14)

  • FAIR (CV)
  • https://scholar.google.com/citations?hl=ko&user=kRJkDakAAAAJ&view_op=list_works&sortby=pubdate

• Natalia Neverova, France (Ph.D 16) (FB 16)

  • FAIR (CV)
  • https://nneverova.github.io/

FAIR, Paris, ML-NLP 6/2/2015

• Holger Schwenk, France (FB 15)

  • FAIR (ML, Translation)
  • https://research.fb.com/people/schwenk-holger/

• Iasonas Kokkinos, France (Ph.D 06) (FB Dec. 16)

  • FAIR (CV)
  • http://www0.cs.ucl.ac.uk/staff/I.Kokkinos/
  • 옮기고나서 ICCV 2편, NIPS17 제출 but reject, CVPR18 1편 제출
  • ocolas Usunier 랑 일을 많이 하네;

• Nocolas Usunier

  • FIAR, NLP

• David Lopez-Paz (ML: Causal Inference), (Ph.D 16), Paris

  • FAIR (ML)
  • https://research.fb.com/people/lopez-paz-david/
  • https://lopezpaz.org/
  • 실적 좋음
  • NIPS 3, ICML 3, JMLR 2, CVPR 1 ICLR 2, Arxiv 2 (ICLR 2018)

• Marc’Aurelio Ranzato (FB13)

  • FAIR, (Dialog Learning)
  • https://research.fb.com/people/ranzato-marcaurelio/

[unizard]

FAIR, NY, ML 12/xx/2013

• Yann LeCun, Director, NYU Professor, Menlo Park, NY, Paris (Ph.D 87) (FB13)

  • FAIR
  • http://yann.lecun.com/

• Jason Weston, NYU Professor, NY (Ph.D 00)

  • FAIR
  • https://research.fb.com/people/weston-jason/
  • Wikipedia를 통해서 스스로 배우는 머신, ML!! Too impressive

• Laurens van der Maaten, DUT Professor, NY (Ph.D 09) (FB15)

  • FAIR
  • http://lvdmaaten.github.io/
  • (CV & ML)

• Benjamin Graham, NY (FB 16)

  • FAIR
  • https://research.fb.com/people/graham-benjamin/
  • (ML, CV)

• Douwe Kiela, NY (Ph.D 16) (FB16)

  • FAIR (Vision, Language)
  • http://www.cl.cam.ac.uk/~dk427/
  • 2017년도 실적;;; 8개? ㅋㅋ 괴물이네;;

• Y-Lan Boureau, NY (Ph.D 15) (FB 15)

  • FAIR (ML,Dialog)
  • http://www.notebleue.org/
  • ML, CV 모두 하는 듯 / 합류하고 ICLR 2017 1편;; 애매한데;

[unizard]

FAIR, Menlo Park

• Piotr Dollar, Research Manager, San Jose (PhD 07), (FB 14)

  • FAIR
  • https://pdollar.github.io/

• Kaiming He, Menlo Park (PhD 11) (FB16)

  • FAIR
  • http://kaiminghe.com/

• Rob Fergus, Professor

  • FAIR

FAIR, Seattle FB ( Detection )

• Ross Girshick, (Ph.D 12), Seattle
  • http://www.rossgirshick.info/

FAIR, Pittsburgh

• Yuandong Tian, (Ph.D13) (FB 15) ==> 자율주행!!
  • https://research.fb.com/people/tian-yuandong/
  • ICLR2017 2, CVPR2017 1, NIPS2017 1 // 한해 평균 4편으로 시작했군.. 잘하고 있네
  • 인턴데리고 잘하고 있음 굿굿
  • Oculus Research 팀이 별도로 피츠버그에서 운영중임. 규모있게 있으나, 사업화쪽이 강함

[unizard]

Applied Machine Learning (ALM) team, Menlo Park

• Du Tran, (Ph.D 16)

  • AML, FAIR, CV
  • http://www.cs.dartmouth.edu/~dutran/
  • 졸업 후 합류해서 CVPR2018 2편 제출해둔 상황임; 모두 1저자;; 너 대단하다. ㅋㅋㅋ
  • 논문 소속이 Facebook 인것을 보아, FAIR 가 아닌듯 한데;;;;

• Guan Pang (CV: 3D computer vision and object detection), (Ph.D 16)

  • AML + CV
  • http://graphics.usc.edu/cgit/people/gpang/index.html
  • 실적이 2nd tier 가 주를 이룸, 아직 페이퍼 없음
  • FB Intern

• Peizhao Zhang, (CV+Mobile Vision Team), (Ph.D 16)

  • AML + CV
  • https://research.fb.com/people/zhang-peizhao/
  • 저널만 3편, ToG 2편, TCSVT 1편,
  • FB Intern

Applied Machine Learning (ALM) team, San Jose (Menlo Park 랑 같은 오피스인가?)

• I. Zeki Yalniz, (Ph.D 13) (FB 17)

  • AML, CV
  • https://research.fb.com/people/yalniz-i-zeki/
  • ECCV 1, TPAMI 1 // X

• Zhicheng Yan (CV: Image & Video Understanding Platform), San Jose, (Ph.D 16)

  • AML, CV
  • https://sites.google.com/view/zcyan
  • TVCG 3, ICMR 2, IJCNN 1, ToG, ICCV, ACL, TIP // Arxiv (CVPR-산학인듯)

• Yan Zhu, (CV), (Ph.D 17)

  • AML, CV
  • https://wakeupbuddy.github.io/www/
  • FB Intern with Dollar, Yuandong Tian
  • 실적 애매함, MICCAI 1, CVIU 1, ICLR 1(인턴), CVPR 1(인턴)

[unizard]

Computer Vision Team + FAIR, Menlo Park

Towards AI agents that can see, talk, and act!

• Dhruv Batra (Visual Dialog), Professor ** (Ph.D 10)

  • CV, FAIR
  • https://research.fb.com/people/batra-dhruv/
  • http://www.cc.gatech.edu/~dbatra/

• Devi Parikh (Visual Dialog), Professor ** (Ph.D 09)

  • CV, FAIR
  • https://research.fb.com/people/parikh-devi/

• Marcus Rohrbach, (Visual Dialog), (Ph.D 14) ==> 이분이 FB 정규직원으로 쭉! 이끌로 가시려나?

  • FAIR, AML, CV
  • http://rohrbach.vision/
  • 이분의 논문 실적은 밑에 정리 했지만, 버클리에서 포닥하면서 논문 제조기였음;;;; 2017 년 메이저 5개;;
  • FB로 넘어온지 얼마 되지 않았고, 지금은 Mohamed Elhoseiny 포닥과 일하는 듯

• Mohamed Elhoseiny (Ph.D 17)

  • FAIR, CV Post Doc
  • https://sites.google.com/site/mhelhoseiny/

Computer Vision Team, Menlo Park

• Fei Yang, Tech Leader (Ph.D 10) (FB 13)

  • CV
  • https://research.fb.com/people/yang-fei/
  • 실적애매하네; 16, 17년 논문 실적이 없음, 리더라서 그런가?

• Albert Gordo, (Ph.D 12)

  • CV, AML
  • Visual Search in XRCE
  • https://agordo.github.io/
  • IJCV1, ECCV1, IJCV1, ICCV1 (1년에 메이저학회1, 저널 1) 그냥 꾸준히 쓰는 스타일 이었다. 제록스에서;;
  • FB 로 온뒤로는 특허만 쓰는걸 봐서; 서비스 하는 듯

• Yannis Kalantidis (Visual Genome) (Ph.D 14)

  • CV, AML
  • Previous work: Visual Search
  • http://www.skamalas.com/#pubs
  • 실적이 workshop이 많군.. 애매함

• Yixuan Li (Ph.D 17) (FB17)

  • CV, AML
  • https://research.fb.com/people/li-yixuan/
  • 실적많은데? FAIR 가 아니네? 이상하군;;;;; 너무 어려서 그런가?
  • WWW 3, ICDM, TKDD, ICLR 2, CVPR //

[unizard]

Computer Vision, Computational Photography Group, Seattle

• Matt Uyttendaele, (MS 89)

  • CV, AML
  • https://research.fb.com/people/uyttendaele-matt/

• Johannes Kopf, (Ph.D 08)

  • CV, AML
  • http://johanneskopf.de/

• Ira Kemelmacher Shlizerman, (Ph.D 09)

  • CV , AML
  • https://homes.cs.washington.edu/~kemelmi/
  • (AR & VR ), ObamaNet

• Kevin Matzen, (Ph.D 15)

  • CV, AML
  • https://research.fb.com/people/matzen-kevin/

[unizard]

FAIR, Satellite team, Israel

• Yaniv Taigman, FB 12

  • FAIR, CV
  • https://ytaigman.github.io/

• Lior Wolf, XProfessor *

  • FAIR
  • https://research.fb.com/people/wolf-lior/

[unizard]

VR, Surround360 team, Melno Park

• Team: https://facebook360.fb.com/facebook-surround-360/
• Albert Parra Pozo (PhD 14) (FB14)
  • CV
  • https://www.linkedin.com/in/albertparra/
  • Research Scientist 인데 최근 논문이 없네?

[unizard]

AI in Games, Game Research, Melo Park

https://github.com/facebookresearch/House3D https://arxiv.org/abs/1801.02209 http://yuandong-tian.com/flair2017.pdf

[unizard]

Marcus Rohrbach, (Visual Dialog), (Ph.D 14) (FB 17)

• [CVPR2018] Memory Aware Synapses: Learning what (not) to forget, FB 쪽으로 넘어오고 시작한 첫 연구임. 자리 잘 잡으시길;;

• [ICCV2017] Learning to Reason: End-to-End Module Networks for Visual Question Answering, Berkeley (포닥하면서 쓴 연구), 3rd

• [ICCV2017] Speaking the Same Language: Matching Machine to Human Captions by Adversarial Training, Berkeley (포닥하면서 쓴 연구), 2nd

• [CVPR2017] Generating Descriptions with Grounded and Co-Referenced People, 2nd

• [CVPR2017] Modeling Relationships in Referential Expressions with Compositional Modular Networks, 2nd

• [CVPR2017] Captioning Images with Diverse Objects, 3rd

• [IJCV2017] [Captioning Images with Diverse Objects]()

• [IJCV2017] [Movie Description]()

[unizard]

Rahaf Aljundi, KU Leuven (관심분야 에이스 발견)

• http://homes.esat.kuleuven.be/~raljundi/index.html
• Recent Works
  • Continual Learning with Hebbian Synapses, CVPR17
  • Encoder Based Lifelong Learning, ICCV17
  • Expert Gate: Lifelong Learning with a Network of Experts, CVPR17

[unizard]

Du Tran (Ph.D 16), (FB16)

• Recent Works

  • [CVPR2018] A Closer Look at Spatiotemporal Convolutions for Action Recognition, 1st
  • [CVPR2018] ConvNet Architecture Search for Spatiotemporal Feature Learning, 1st

• 애쓰고 있네;;; 인턴을 대리고 했는지는 저자를 좀 더 분석 해봐야겠지만;;;;;

[unizard]

Mohamed Elhoseiny (Ph.D 17) PostDoC

• Recent Works
  • [CVPR2018] Imagine it for me: Generative Adversarial Approach for Zero-Shot Learning from Noisy Texts 학교에서 하던 연구
  • [CVPR2018] Memory Aware Synapses: Learning what (not) to forget, FB로 넘어와서 한 연구;;;

[unizard]

Douwe Kiela, NY (Ph.D 16) (FB16)

• NLP+VIsion 즉, 영상을 말로 잘 설명하거나 하는 연구;;
• Recent Works
  • [AAAI2018] Efficient Large-Scale Multi-Modal Classification, D. Kiela, E. Grave, A. Joulin and T. Mikolov, 그림과 해쉬대그를 이용한 분류기를 만드는 것 같은데;;;;
  • [ICLR2018] MASTERING THE DUNGEON: GROUNDED LANGUAGE LEARNING BY MECHANICAL TURKER DESCENT All FB
• [NIPS2017] Poincaré Embeddings for Learning Hierarchical Representations All FB
• Learning Visually Grounded Sentence Representations , All FB

[unizard]

Big Picture of Facebook Research

Computer Vision Team + FAIR, Menlo Park

(1) Towards AI agents that can see, talk, and act! (Talk!!) ==> "NAVER 지식인" (only based on fact without subjective information such as recommendation)

FAIR, Menlo Park, Seattle

Towards Human Level Vision (Best of Best..See) ==> Core tech of (1)

FAIR, NY

Towards Human Level Language (Best of Best... Talk) ==> Core tech of (1)

FAIR, Pittsburgh

They do not consider simple Game Agent. We can start to understand the real world with interaction. Namely, this is the abstract of the real world.

• http://yuandong-tian.com/flair2017.pdf

  Towards AI agents that can see and act! Towards Human Level Act (See는 깔고 가는 듯) (Abstract game to real-world)

???

Towards Human Level Learning (Continual Learning) ==> "The Ultimate Jarvis"

[unizard]

NELL, NEIL 과 같은 연구를 하고 있는가? (해당연구는 Ai2에서 하고 있는건가?)

아니면, web scale data를 핸들링하기에 쉽지 안아 포기 한건가?

• Instagram의 image-hashtag pairs
• wikipedia의 Word-meaning pairs
• 지식인의 Question-Best Answer pairs

[unizard]

• StarSpace: Embed All The Things! AAAI18

• Efficient Large-Scale Multi-Modal Classification, AAAI18

• One-Sided Unsupervised Domain Mapping, NIPS17

• ELF: An Extensive, Lightweight and Flexible Research Platform for Real-time Strategy Games, NIPS2017

• Unbounded Cache Model for Online Language Modeling with Open Vocabulary, NIPS2017

• VAIN: Attentional Multi-agent Predictive Modeling, NIPS2017

• Gradient Episodic Memory for Continual Learning, NIPS2017

• Poincaré Embeddings for Learning Hierarchical Representations, NIPS2017

• Attentive Explanations: Justifying Decisions and Pointing to the Evidence, NIPS2017

• On the Optimization Landscape of Tensor Decompositions, NIPS2017

• Best of Both Worlds: Transferring Knowledge from Discriminative Learning to a Generative Visual Dialog Model, NIPS2017

• Fader Networks: Manipulating Images by Sliding Attributes, NIPS2017 (!!)

• Casual 3D Photography, SIG Asia 2017 (!!) Oculus 에서 쓰려고 만들었나? 핸드폰으로 스몰모션으로 촬영하고 나서, 이를 나중에 집에서 3D로 보고 싶은 연구임;;;

• Bringing Portraits to Life, Sig Asia 2017