I am a Research Scientist at Meta where I work with the Instagram ranking team. My long term goal is to build algorithms that can continuously learn from humans while coexisting with them in shared spaces. In particular I am interested in neural nets and unsupervised learning of latent attributes. I graduated from the MScAC program at UofT Computer Science under the supervision of Dr. Gerald Penn, Jithin Pradeep and Hussain Zaidi.
I completed my BEng. in CS from RV College of Engineering, Bangalore in 2018. Outside of work, I'm an amateur astronomer, flautist and an avid reader. Do you love space too? Checkout my bookshelf here!
Activity
[last updated July 2022]
- June 2022: Joined Facebook Toronto as a research scientist.
- Apr 2022: Preprint of our work 'Generalized Distribution Calibration for Few-Shot Learning' is now available on arxiv.
-
Apr 2022: Invited talk on our Few-Shot Learning research at McGill University.
Recordings here [ YouTube ] -
Aug 2021: Presenting our work on latent representations at IWSLT2021.
Check the recording here [ YouTube ] - June 2021: On MScAC Grad21 Spotlight [ News ]
-
March 2021: Helping with a Cause– I have partnered with mentr-me.com to help prospective graduate students and ML aspirants for a small donation. Find out more here–
mentr-me.com/user/shakti-kumar - Feb 2021: Our paper on Learning Latent Representation in Speech is now accepted at ACL2021.
- Dec 2020: Student Presentation Award, ARIA 2020, Dept. of Computer Science [ News ]
-
June 2019: Received the Vector Scholarship in Artificial Intelligence 2019, $17,500 [ News ]
-
Received Summer Research Fellowship Award 2016 [ News ]