K
Kenneth Marino
Researcher at Carnegie Mellon University
Publications - 19
Citations - 1347
Kenneth Marino is an academic researcher from Carnegie Mellon University. The author has contributed to research in topics: Computer science & Reinforcement learning. The author has an hindex of 7, co-authored 14 publications receiving 787 citations. Previous affiliations of Kenneth Marino include Facebook.
Papers
More filters
Proceedings ArticleDOI
The Pose Knows: Video Forecasting by Generating Pose Futures
TL;DR: This work exploits human pose detectors as a free source of supervision and breaks the video forecasting problem into two discrete steps, and uses the structured space of pose as an intermediate representation to sidestep the problems that GANs have in generating video pixels directly.
Proceedings ArticleDOI
OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledge
TL;DR: Recently, this paper proposed the OK-VQA dataset, which includes more than 14,000 questions that require external knowledge to answer and showed that the performance of state-of-the-art VQA models degrades drastically in this new setting.
Proceedings ArticleDOI
The More You Know: Using Knowledge Graphs for Image Classification
TL;DR: This paper investigates the use of structured prior knowledge in the form of knowledge graphs and shows that using this knowledge improves performance on image classification, and introduces the Graph Search Neural Network as a way of efficiently incorporating large knowledge graphs into a vision classification pipeline.
Posted Content
The More You Know: Using Knowledge Graphs for Image Classification
TL;DR: Graph Search Neural Networks (GSNNs) as discussed by the authors use structured prior knowledge in the form of knowledge graphs and show that using this knowledge improves performance on image classification, which is similar to our work.
Proceedings ArticleDOI
KRISP: Integrating Implicit and Symbolic Knowledge for Open-Domain Knowledge-Based VQA
TL;DR: Knowledge Reasoning with Implicit and Symbolic rePresentations (KRISP) as mentioned in this paper combines implicit knowledge from unsupervised language pre-training and symbolic knowledge encoded in knowledge bases.