Výsledky vyhledávání - "Veerabadran P"

Report

Adaptive recurrent vision performs zero-shot computation scaling to unseen difficulty levels

Autor: Veerabadran, Vijay, Ravishankar, Srinivas, Tang, Yuan, Raina, Ritik, de Sa, Virginia R.

Humans solving algorithmic (or) reasoning problems typically exhibit solution times that grow as a function of problem difficulty. Adaptive recurrent neural networks have been shown to exhibit this property for various language-processing tasks. Howe

Externí odkaz: http://arxiv.org/abs/2311.06964

Zobrazit plný text záznamu

Akademický článek

Homing of translocated native Indian pythons in Moyar River Valley, South India

Autor: C.S. Vishnu, Chinnasamy Ramesh, Shannon Pittman, Vedagiri Thirumurugan, Gautam Talukdar, Krishnamurthy Ashokan, P.G. Arunlal, Veerabadran Naganathan

Publikováno v: Heliyon, Vol 10, Iss 14, Pp e33010- (2024)

Homing is often a critical aspect of an animal's behavioural and spatial ecology. Translocation is considered to be a wildlife management strategy that could reduce human-wildlife confrontation, but this strategy may not be effective if animals attem

Externí odkaz: https://doaj.org/article/7dcd02102e164321bc232f78c5ae1f5c

Zobrazit plný text záznamu

Akademický článek

Subtle adversarial image manipulations influence both human and machine perception

Autor: Vijay Veerabadran, Josh Goldman, Shreya Shankar, Brian Cheung, Nicolas Papernot, Alexey Kurakin, Ian Goodfellow, Jonathon Shlens, Jascha Sohl-Dickstein, Michael C. Mozer, Gamaleldin F. Elsayed

Publikováno v: Nature Communications, Vol 14, Iss 1, Pp 1-12 (2023)

Abstract Although artificial neural networks (ANNs) were inspired by the brain, ANNs exhibit a brittleness not generally observed in human perception. One shortcoming of ANNs is their susceptibility to adversarial perturbations—subtle modulations o

Externí odkaz: https://doaj.org/article/e4a0c04dd90842a19effeed01013b872

Zobrazit plný text záznamu

Report

Learning compact generalizable neural representations supporting perceptual grouping

Autor: Veerabadran, Vijay, de Sa, Virginia R.

Work at the intersection of vision science and deep learning is starting to explore the efficacy of deep convolutional networks (DCNs) and recurrent networks in solving perceptual grouping problems that underlie primate visual recognition and segment

Externí odkaz: http://arxiv.org/abs/2006.11716

Zobrazit plný text záznamu

Report

Adversarial Distortion for Learned Video Compression

Autor: Veerabadran, Vijay, Pourreza, Reza, Habibian, Amirhossein, Cohen, Taco

In this paper, we present a novel adversarial lossy video compression model. At extremely low bit-rates, standard video coding schemes suffer from unpleasant reconstruction artifacts such as blocking, ringing etc. Existing learned neural approaches t

Externí odkaz: http://arxiv.org/abs/2004.09508

Zobrazit plný text záznamu

Report

Learning long-range spatial dependencies with horizontal gated-recurrent units

Autor: Linsley, Drew, Kim, Junkyung, Veerabadran, Vijay, Serre, Thomas

Progress in deep learning has spawned great successes in many engineering applications. As a prime example, convolutional neural networks, a type of feedforward neural networks, are now approaching -- and sometimes even surpassing -- human accuracy o

Externí odkaz: http://arxiv.org/abs/1805.08315

Zobrazit plný text záznamu