Search Results for author:
Found 2 papers, 2 papers with code
Understanding the Effects of Second-Order Approximations in Natural Policy Gradient Reinforcement Learning
Natural policy gradient methods are popular reinforcement learning methods that improve the stability of policy gradient methods by utilizing second-order approximations to precondition the gradient with the inverse of the Fisher-information matrix.
M2A: Motion Aware Attention for Accurate Video Action Recognition
We further compared the performance of M2A with other state-of-the-art motion and attention mechanisms on the Something-Something V1 video action recognition benchmark.
Cannot find the paper you are looking for? You can
Submit a new open access paper.
