Gemmechu Hassena

I am a CS PhD student at Cornell University working in the intersection of ML, CV and Graphics, advised by Bharath Hariharan and Steve Marschner. Prior to Cornell I earned my Bachelor's in Software Engineering from Addis Ababa University, Ethiopia. I have worked as a research intern at INRIA, Ecole Polytechnique with Maks Ovsjanikov. and at Univeristy of Michigan with David Fouhey on a Computer vision, Graphics and machine learning projects.

Email  /  CV  /  Linkedin  /  Twitter  /  Github

[May 2024] Awarded two awards "Exemplary Service Award for Early Career Graduate Students" and "Bowers CIS Distinguished Leadership in Service Award"

[April 2024] Presented poster at NYC Computer Vision Day, NYU

[Dec 2023] Gave a talk at DevFest Addis 2023

[Dec 2023] Gave Lightning talk at NYC vision workshop, Cornell Tech

[August 2022] Started my PhD at Cornell!

[March 2022] Started working as a Team lead in Product Management at A2SV/Eskalate

[Oct 2021] Started working as a research intern at INRIA, Ecole Polytechnique

[Oct 2021] Graduated from Addis Ababa University, bachelor of science in Software Engineering with Great distinction!

[Jun 2021] Quantifying Bird Skeletons accepted at CV4Animals Workshop In conjunction with CVPR 2021 [paper]

[Apr 2021] published a blog post on University of Michigan AI website! [blog]

[May 2020] Started working as a research intern at University of Michigan

[Dec 2019] Gave a talk on Afro-FM radio station about INSPIRE student club, AI in Ethiopia conference and the ICLR conference to be held in Addis Ababa

[Nov 2019] Organized Twitter CEO Jack Dorsey's visit to Addis Ababa University [photo]

[Nov 2019] Invited Tensorflow Speaker on Devfest 2019 - Google Developer Group Addis [website]

[Oct 2019] Organized the 1st AI in Ethiopia Conference [website]

How can we separate objects in a scene without a segmentation mask? Given multi-view images and point clicks in a single view, ObjectCarver will generate a mask, propagate it to other views, reconstruct the 3D surface, and separate the individual objects in the scene.

We introduced a dataset and system for measuring skeletal traits from museum specimens, which speeds up skeleton measurement by 15x. This research helped Ecologists in their study of how global warming causes the size of birds to shrink

Monocular depth estimation, we developed a model that estimates depth and floor of a scene from RGB image, by using people's height as the reference.

Developed a 3D game with Unity engine

[Jun 2024] Research Overview, Host: Dr. LeeAnn Roberts, SoNiC Cornell

[Apr 2024] Lightining talk Grad Cohort for IDEALS, Minneapolis

[Mar 2024] Research talk, Host: Prof. Maks Ovsjanikov, Ecole Polytechnique

[Mar 2024] Research talk, Host: Prof. Abe Davis, Vision and Graphics Seminar, Cornell

[Dec 2023] Title: 3D for Everyone co-presented with Milky Hassena, Devfest-addis 2023 Addis Ababa

[Dec 2023] Lightning paper presentation, NYC vision workshop at Cornell Tech

Template