Dhruv Batra

TL;DR: These approaches, based on LSTM-RNNs, VQA model uncertainty, and caption-question similarity, are able to outperform strong baselines on both relevance tasks and are shown to be more intelligent, reasonable, and human-like than previous approaches.

TL;DR: EvalAI is built to provide a scalable solution to the research community to fulfill the critical need of evaluating machine learning models and agents acting in an environment against annotations or with a human-in-the-loop.

TL;DR: SplitNet is proposed, a method for decoupling visual perception and policy learning by incorporating auxiliary tasks and selective learning of portions of the model that explicitly decompose the learning objectives for visual navigation into perceiving the world and acting on that perception.

TL;DR: This work develops the first approximate inference algorithm for 1-Best (and M-Best) decoding in bidirectional neural sequence models by extending Beam Search to reason about both forward and backward time dependencies, and introduces a novel Fill-in-the-Blank Image Captioning task which requires reasoning about both past and future sentence structure to reconstruct sensible image descriptions.

TL;DR: This paper developed a language-guided navigation task set in a continuous 3D environment where agents must execute low-level actions to follow natural language navigation directions, which lifted a number of assumptions implicit in prior work that represents environments as a sparse graph of panoramas with edges corresponding to navigability.