Zobrazeno 1 - 3
of 3
pro vyhledávání: '"Seshadri, Amrit Diggavi"'
Autor:
Seshadri, Amrit Diggavi
With the size and cost of large transformer-based language models growing, recently, there has been interest in shortcut casting of early transformer hidden-representations to final-representations for cheaper model inference. In particular, shortcut
Externí odkaz:
http://arxiv.org/abs/2409.14091
In this work, following the intuition that adverbs describing scene-sequences are best identified by reasoning over high-level concepts of object-behavior, we propose the design of a new framework that reasons over object-behaviours extracted from ra
Externí odkaz:
http://arxiv.org/abs/2307.04132
Synthesizing high-quality, realistic images from text-descriptions is a challenging task, and current methods synthesize images from text in a multi-stage manner, typically by first generating a rough initial image and then refining image details at
Externí odkaz:
http://arxiv.org/abs/2110.08143