No abstract

No abstract

Machine Learning can work very well with image recognition, but it is used to recognize audio patterns. Machine listening identifies audio patterns of different entities like the car engine, human speaking, nature sounds, etc. The environmental sound classification plays an important role to encourage citizens to travel smartly within a city without creating unbearable noises. On the other hand, it also promotes the city council to maintain and predict a sustainable sound at rush hour with ins the city. The aim of this early-stage research is to present a methodology that will read the labeled audio files, extract features from them, feed features to a sequential model. Moreover, the model will have the ability to classify these audio files of vehicles based on their input feature(s) and then further categorize them as it either light-weight, medium-weight, heavy-weight, rail-bound or two-wheeled vehicle using the applications of machine listening and deep learning in the field of sound acoustics. Therefore, It will also classify unlabelled test data files on a pre-trained model. This research provides us the base model for the vehicle classification giving both advantages and disadvantages along with the possibility for future extensions. CCS CONCEPTS• Software and its engineering → Integrated and visual development environments; • General and reference → Performance; • Computing methodologies → Concurrent computing methodologies; Massively parallel algorithms.

Electromobility has profound economic and ecological impacts on human society. Much of the mobility sector’s transformation is catalyzed by digitalization, enabling many stakeholders, such as vehicle users and infrastructure owners, to interact with each other in real time. This article presents a new concept based on deep reinforcement learning to optimize agent interactions and decision-making in a smart mobility ecosystem. The algorithm performs context-aware, constrained optimization that fulfills on-demand requests from each agent. The algorithm can learn from the surrounding environment until the agent interactions reach an optimal equilibrium point in a given context. The methodology implements an automatic template-based approach via a continuous integration and delivery (CI/CD) framework using a GitLab runner and transfers highly computationally intensive tasks over a high-performance computing cluster automatically without manual intervention.

No abstract