Ishan Banerjee

Stephan Arlt, Ishan Banerjee, Cristiano Bertolini et al.

Graphical user interfaces (GUIs), due to their event driven nature, present a potentially unbounded space of all possible ways to interact with software. During testing it becomes necessary to effectively sample this space. In this paper we develop algorithms that sample the GUI's input space by only generating sequences that (1) are allowed by the GUI's structure, and (2) chain together only those events that have data dependencies between their event handlers. We create a new abstraction, called an event-dependency graph (EDG) of the GUI, that captures data dependencies between event handler code. We develop a mapping between EDGs and an existing black-box user-level model of the GUI's workflow, called an event-flow graph (EFG). We have implemented automated EDG construction in a tool that analyzes the bytecode of each event handler. We evaluate our "grey-box" approach using four open-source applications and compare it with the current state-of-the-art EFG approach. Our results show that using the EDG reduces the number of test cases while still achieving at least the same coverage. Furthermore, we were able to detect 2 new bugs in the subject applications.

Priyaranjan Pattnayak, Hitesh Laxmichand Patel, Bhargava Kumar et al.

Multimodal learning, a rapidly evolving field in artificial intelligence, seeks to construct more versatile and robust systems by integrating and analyzing diverse types of data, including text, images, audio, and video. Inspired by the human ability to assimilate information through many senses, this method enables applications such as text-to-video conversion, visual question answering, and image captioning. Recent developments in datasets that support multimodal language models (MLLMs) are highlighted in this overview. Large-scale multimodal datasets are essential because they allow for thorough testing and training of these models. With an emphasis on their contributions to the discipline, the study examines a variety of datasets, including those for training, domain-specific tasks, and real-world applications. It also emphasizes how crucial benchmark datasets are for assessing models' performance in a range of scenarios, scalability, and applicability. Since multimodal learning is always changing, overcoming these obstacles will help AI research and applications reach new heights.