We address a question answering task on real-world im- ages that is set up as a Visual Turing Test. By combining latest advances in image representation and natural language processing, we propose Neural-Image-QA, an end-to-end formulation of this problem for which all parts are trained jointly. In contrast to previous efforts, we are facing a multi-modal problem where the language output (answer) is conditioned on visual and natural language input (ques- tion). Our result doubles the performance of the previous best result on this problem. We provide additional insights into the problem by analyzing how much information is con- tained only in the language part for which we provide a new human baseline. Further annotations were collected to study human consensus, which is related to the ambiguities inherent in this challenging task.
from cs.AI updates on arXiv.org http://ift.tt/1GZzcre
via IFTTT
No comments:
Post a Comment