Hi, this is
I am a PhD student at Johns Hopkins University, where I am affiliated with the Mathematical Institute
for Data Science and have been extremely fortunate to be advised by
Prof. René Vidal and Prof. Joshua Vogelstein.
I am also an alumni of Erasmus Mundus Master's Joint Program in Image Processing and
Computer Vision, where I was very grateful to work as a Research Assistant at VPU Lab of the Autonomous
University of Madrid and as a Scientific Employee at TU Graz Institute of Computer Graphics
and Vision to do my Master’s thesis on ‘Lifelong learning on Autonomous Vehicles’ under the supervision of Prof.
My research interests are in Image Processing, Computer Vision, and Machine Learning.
I'm extremely enthusiastic about AI and its power to make our world a better place to live; and I want to contribute my part on this captivating and wonderful scientific field working to solve the open problems in Computer Vision and Machine Learning so as to build autonomous systems that can "see" and reason.
[Jul 12, 2023] Our workshop and competition on 'LLM Efficiency Challenge: 1 LLM + 1GPU + 1Day' has been accepted to NeurIPS 2023.
[May 22, 2023] I have started a new position as a PhD Reident @ X (formerly Google X).
[May 31, 2022] I have started a new position as a Research Intern at Microsoft Research.
[Sep 16, 2021] I have been accepted to Google’s CSRMP Class of 2021B.
[Aug 16, 2021] I have joined The Africa I Know as a Contributing Editor.
[Jul 23, 2021] Our workshop on 'Out-of-distribution generalization and adaptation in natural and artificial intelligence' has been accepted to NeurIPS 2021. I am honoured to serve on the Programme Committee.
Bachelor's Thesis Title: 'An Autonomous Traffic Management System'
Kaleab is a highly motivated researcher with an outstanding academic profile, who is able to rapidly embed himself into the thick of any problem and learn quickly and work independently. I found him to be a bright and engaging member of our lab with an optimistic and energetic attitude.Prof. Joshua Vogelstein - director of the NeuroData lab at Johns Hopkins University.
Mr. Kinfu worked on state-of-the-art lifelong learning and unsupervised domain adaptation techniques for autonomous vehicles under my supervision. He demonstrated the ability to work independently with great creativity and enthusiasm. He then wrote his thesis, which is clear, concise, and interesting to read. He proved to be a capable and motivated researcher with all the necessary qualities.Prof. Dr. Horst Bischof - Rector @ TU Graz, Treasurer @ European Computer Vision Association.
I supervised Mr. Kinfu on a research project entitled '3D Shape Prediction from a single image using deep learning'. The project consist of the realization of a state-of-the-art method. He was very participating and he showed great competences to work in the topic. His qualities made his project a very successful one.Prof. Pascal Desbarats - Directeur adjoint EDMI de la spécialité Informatique. Resp. Master informatique parcours IPCV, France
Mr. Kinfu attended my classes at PPCU, which he fulfilled with an excellent grade, moreover he was among the few students who obtained an offered mark. He was also exceptionally motivated and was successful in getting Excellent marks in all other courses. I also had the opportunity to check his research works and I can confirm that they are excellent contributions to the very competitive R&D field.Dr Csaba Benedek - Associate Professor @ Pazmany Peter Catholic University, Hungary.
Kaleab has demonstrated very good programming skills, including the ability to work with different programming frameworks and code versioning tools. He has shown a high level of organization, and capabilities to work in a research team, actively proposing creative solutions in our weekly follow-up meetings.Dr. Pablo Carballeira Lopez - Assistant Professor @ Autonomous University of Madrid, Spain.
Mr. Kinfu displayed an outstanding analytical skills and a strong ability to grasp complex ideas with ease. He is a skilful programmer with strong initiative. He proved to be diligent and systematic and was able to solve any given task in the scheduled time. Mr. Kinfu obtained deep insights on selected state-of-the-art image and video analysis techniques and developed effective research abilities.Dr Marcos Escudero Viñolo - Assistant Professor @ Autonomous University of Madrid, Spain.
Over the years, I have had the pleasure of watching this young man blossom into a focused, self-assured scholar who aspires to be a great innovator and researcher.Amsale Zelalem, PhD Candidate @ University of Twente, Netherlands & Lecturer @ Addis Ababa University, Ethiopia
Kaleab showed innovation, excellent programming skills, and good skills of market needs assessment to develop an application that has very good promise.Tewodros Mebrahtu, Director @ Ethiopian ICT Center of Excellence
Extensions of Adversarial Training for Video Classification
Lifelong Learning for Autonomous Vehicles
Multi-task learning of Monocular Depth Estimation and Segmentation from Video
3D Shape Prediction from Single RGB Image using Deep Learning
Ground-truth generation for Google Street View images
Ethiopian Sign Language to Amharic Text Translator
Automated Optical Mark Reader
Intelligent Traffic Management System
Multiple Object Recognition using Deep Learning
Abugida: Educational App for Preschoolers
Face Recognition | Body Tracking