A
Akhil S. Raju
Researcher at Massachusetts Institute of Technology
Publications - 4
Citations - 292
Akhil S. Raju is an academic researcher from Massachusetts Institute of Technology. The author has contributed to research in topics: Computer science. The author has an hindex of 1, co-authored 1 publications receiving 238 citations.
Papers
More filters
Proceedings ArticleDOI
Understanding and Predicting Image Memorability at a Large Scale
TL;DR: LaMem is built, the largest annotated image memorability dataset to date, using Convolutional Neural Networks, to demonstrate that one can now robustly estimate the memorability of images from many different classes, positioning memorability and deep memorability features as prime candidates to estimate the utility of information for cognitive systems.
Journal ArticleDOI
Vision-Language Models as Success Detectors
Yuqing Du,Ksenia Konyushkova,Misha Denil,Akhil S. Raju,Jessica Landon,Felix Hill,Nando de Freitas,Serkan Cabi +7 more
TL;DR: In this paper , the authors treat success detection as a visual question answering (VQA) problem and investigate the generalisation properties of a Flamingo-based success detection model across unseen language and visual changes in the first two domains.
Journal ArticleDOI
RoboCat: A Self-Improving Foundation Agent for Robotic Manipulation
Konstantinos Bousmalis,Giulia Vezzani,Dushyant Rao,Coline Devin,Alex Xavier Lee,Maria Bauza,Todor Davchev,Yuxiang Zhou,Agrim Gupta,Akhil S. Raju,Antoine Laurens,Claudio Fantacci,Valentin Dalibard,Martina Zambelli,Murilo Fernandes Martins,Rugile Pevceviciute,M. L. Blokzijl,Misha Denil,Nathan Batchelor,Thomas Lampe,Emilio Parisotto,Konrad Żołna,Scott Reed,Sergio Gomez Colmenarejo,Jon Scholz,Abbas Abdolmaleki,Oliver Groth,J.-B. Regli,Oleg O. Sushkov,José Enrique Chen,Yusuf Aytar,David Barker,Martin Riedmiller,Jost Tobias Springenberg,Raia Hadsell,Francesco Nori,Nicolas Heess +36 more
TL;DR: In this article , a visual goal-conditioned decision transformer capable of consuming multi-embodiment action-labeled visual experience is proposed for robotic manipulation, which can generalise to new tasks and robots, both zero-shot as well as through adaptation.
Proceedings ArticleDOI
Automatic Visual Inspection - Defects Detection using CNN
TL;DR: In this article , the PyTorch pipeline is used to divide images into good and anomalous classes and, if the image is categorized as an anomaly, a bounding box is returned for the fault.