In this internship, we aim to develop a machine-learning model that learns human pose and movements. Such
models can be used for AI-assisted human animation creation, motion refinement and denoising, video or imageto-
pose conversion, and motion editing. We plan to implement and push state-of-the-art human motion models by
taking inspiration from the advances in computer vision research.

Cybersecurity incidents such as data breaches have become more critical and costly than ever, as we are generating, processing, and relying on more digital information every day. To quickly identify potential security attacks and prevent them in a sea of system activities, machine learning (ML) has been applied to support security analysts’ decisions. However, the task is challenging as cybersecurity data is often large, noisy, and complex, and there lack sufficient ground truth. Further, ML models are often a black box to the analysts, which usually produce results not easily interpretable.

With the evolution of computer vision methods in recent years, more and more scene-understanding methods have been proposed. Meanwhile, SOTI also designed and developed several scene-understanding algorithms. However, due to the limited computing capabilities of UAVs, all these algorithms need to be compressed to use fewer resources. This project will deliver a model compression and deployment pipeline.

The objective of this project is to develop a software solution that can analyze the accelerometer data on android devices and report certain metrics of a moving vehicle related to road safety and driving conditions. The reports can then be used for immediate call for action in case of detecting an emergency or collecting the statistical data over a longer time period for assessing the general driving behavior.

The proposed research project aims to improve the way data is transmitted between mobile devices and company servers. This will improve the speed and security of file transfers, data synchronization, application deployment, and distant management of mobile devices. The intern will collaborate with experts from the partner organization who will offer guidance and support in this research to identify new techniques that can be used to reduce the amount of data being transmitted, while also ensuring that the data is secure.

Mobile devices have become a crucial tool for businesses, and SOTI MobiControl is a leading mobile device management solution that provides remote control capabilities. However, to ensure proper product functionality and scalability of SOTI MobiControl, the company is looking to research the simulation of remote controlling a mobile device for automation testing. By testing the remote-control feature under various scenarios and conditions, SOTI can identify and address any issues that may arise, resulting in a better-performing product and improved user experience.

The proposed project seeks to develop a Machine Learning-based software solution that accurately measures the capacity, State of Health (SoH), State of Charge (SoC), and cycle count of non-smart batteries utilized in mobile fleets. The project's primary objective is to bridge the gap between smart and non-smart batteries by monitoring non-smart battery capacity and other pertinent parameters.

The proposed research project will focus on analyzing and optimizing the cloud infrastructure used by SOTI to manage mobile devices globally. The intern will analyze the current cloud architecture and hosting costs, identify areas for improvement, and propose and implement optimizations to reduce system requirements and minimize costs. The expected benefit to SOTI is a more cost-effective and efficient cloud infrastructure that maintains the performance and quality of their technology solutions.

Continuous integration (CI) and continuous delivery (CD) are practices that help software development teams deliver code changes more often and with fewer issues. To ensure that code changes are working as they should, developers use Behavior Driven Development (BDD) tests. But running all these tests against every code change can be time-consuming and costly. This project aims at classifying and categorizing the BDD tests into smaller categories and creating a recommendation system that assigns the right tests to each code change.

Investigative journalism is understood to be vital in a democracy. However, such work has long been difficult to sustain through traditional commercial news business models and an ongoing crisis facing these models has greatly diminished the resources available for it at many media outlets. Moreover, traditional forms of investigative journalism have been faulted for not sufficiently engaging diverse audiences and a lack of diversity among practitioners has also been identified as a major issue.