Abstract: Accuracy of transcript is of foremost importance in Automatic Speech Recognition (ASR). State of the art system mostly rely on spelling correction based contextual improvement in ASR, which is generally a static vocabulary based biasing approach. Embodiments of the present disclosure provide a method and system for visual context aware ASR. The method provides biasing using shallow fusion biasing approach with a modified beam search decoding technique, which introduces a non-greedy pruning strategy to allow biasing at the sub-word level. The biasing algorithm brings in the visual context of the robot to the speech recognizer based on a dynamic biasing vocabulary, improving the transcription accuracy. The dynamic biasing vocabulary, comprising objects in a current environment accompanied by their self and relational attributes, is generated using a bias prediction network that explicitly adds label to objects, which are detected and captioned via a state of the art dense image captioning network.

Abstract: This disclosure relates generally to system and methods for dynamic scheduling and rescheduling using heterogeneous multi-agent fleet. The embodiments of present disclosure herein address unresolved problem of task allocation using a single solution which is not sufficient for handling various scenarios of multi-agent task allocation problems. For instance, the task allocation becomes more challenging in a scenario where each task has a deadline associated with it and execution time of the tasks cannot be pre-computed as they are dependent on previous tasks. The method of present disclosure provides a scalable solution for dynamic scheduling and rescheduling that handles tasks with multiple pickup and drop locations and dynamic execution time using the agents with heterogeneous speed in a more efficient manner, reducing indirect operating costs and increasing revenue potential while minimizing additional penalty due to run time delays that an agent may encounter.