Abstract: Methods, apparatus and systems are provided for training a first reinforcement-learning (RL) agent and a second RL agent coupled to a computer game environment using RL techniques. The first RL agent iteratively generates a sub-goal sequence in relation to an overall goal within the computer game environment, where the first RL agent generates a new sub-goal for the sub-goal sequence after a second RL agent, interacting with the computer game environment, successfully achieves a current sub-goal in the sub-goal sequence. The second RL agent iteratively interacts with the computer game environment to achieve the current sub-goal in which each iterative interaction includes an attempt by the second RL agent for interacting with the computer game environment to achieve the current sub-goal. The first RL agent is updated using a first reward issued when the second RL agent successfully achieves the current sub-goal.

Abstract: Systems and methods of curiosity driven reinforcement learning agents promote novel exploration of a virtual interactive environment. The data of the exploration can be stored in a buffer to determine, generate, and display visualizations in the virtual interactive environment. The visualizations can correspond to identify issues in the virtual interactive environment and/or identify relationships between regions of the virtual interactive environment.

Abstract: Methods, apparatus and systems are provided for training a first reinforcement-learning (RL) agent and a second RL agent coupled to a computer game environment using RL techniques. The first RL agent iteratively generates a sub-goal sequence in relation to an overall goal within the computer game environment, where the first RL agent generates a new sub-goal for the sub-goal sequence after a second RL agent, interacting with the computer game environment, successfully achieves a current sub-goal in the sub-goal sequence. The second RL agent iteratively interacts with the computer game environment to achieve the current sub-goal in which each iterative interaction includes an attempt by the second RL agent for interacting with the computer game environment to achieve the current sub-goal. The first RL agent is updated using a first reward issued when the second RL agent successfully achieves the current sub-goal.

Abstract: Systems and methods of curiosity driven reinforcement learning agents promote novel exploration of a virtual interactive environment. The data of the exploration can be stored in a buffer to determine, generate, and display visualizations in the virtual interactive environment. The visualizations can correspond to identify issues in the virtual interactive environment and/or identify relationships between regions of the virtual interactive environment.

Abstract: Methods, apparatus and systems are provided for training a first reinforcement-learning (RL) agent and a second RL agent coupled to a computer game environment using RL techniques. The first RL agent iteratively generates a sub-goal sequence in relation to an overall goal within the computer game environment, where the first RL agent generates a new sub-goal for the sub-goal sequence after a second RL agent, interacting with the computer game environment, successfully achieves a current sub-goal in the sub-goal sequence. The second RL agent iteratively interacts with the computer game environment to achieve the current sub-goal in which each iterative interaction includes an attempt by the second RL agent for interacting with the computer game environment to achieve the current sub-goal. The first RL agent is updated using a first reward issued when the second RL agent successfully achieves the current sub-goal.

Abstract: The present disclosure provides a system for automating graphical testing during video game development. The system can use Deep Convolutional Neural Networks (DCNNs) to create a model to detect graphical glitches in video games. The system can use an image, a video game frame, as input to be classified into one of defined number of classifications. The classifications can include a normal image and one of a plurality of different kinds of glitches. In some embodiments, the glitches can include corrupted textures, including low resolution textures and stretched textures, missing textures, and placeholder textures. The system can apply a confidence measure to the analysis to help reduce the number of false positives.