Abstract: Though premium GPUs sport a high amount of RAM memory and have a large number of cores, in many cases the captured models from a CT scan or MRI need to be inspected in medium to low-end devices, with average GPUs. Under these circumstances, reducing the size of the original dataset while still maintaining the quality is challenging. Most of the previous approaches concentrate on reducing the dataset by using compression techniques that require enormous horsepower to be rendered in realtime. Commercial software, on the other hand, typically only reduces the dataset without any specific technique to maintain data quality. In this talk, I will present some techniques tailored to avoid the many problems that arise when reducing datasets, such as the loss of details, or the changes in color.
We invite all fellow researchers and students to enter into the discussion.
