Recently, high-performing models were developed for image processing and computer vision tasks such as design education, medical diagnosis, precision livestock farming, and weather forecasting, to name but a few. However, enough research has not been conducted on techniques for assessing and evaluating these models with benchmark datasets on transdisciplinary visual communication. Data, whether from primary or secondary sources, in their different formats such as texts, images, videos, etc., are essential materials for conducting scientific experiments in various fields for different purposes. Therefore, in this paper, we discuss the different techniques and models for processing scientific data. The assessment and evaluation of these techniques including their performance are also discussed. This paper is an insight into transdisciplinary visual communication as a way of thinking and working that combines knowledge from different disciplines to create new solutions to complex problems.