Choose timezone
Your profile timezone:
Image Processing Meeting (FTK-IAPP)
→
Europe/Rome
241 (Edificio C)
241
Edificio C
Minutes - Calliope
Status report from fw:
Pierluigi: Data loading to the AM chip done, script for testing missing. Started development of the chip readout
Calliope: Training module control almost finished, missing selected patterns readout for control
Statur report from sw:
Akis ready to produce data for bit-accurate simulation.
Mario occupied with 3D processing (please also see Mario’s slides):
To do list:
1. Proof of concept: use a sample provided by Alessandra with the proper resolution (this still remains a question) to produce filtered images with the same arbitrary luminance thresholds (make a better resolution plot for the threshold selection). We will use one data set to train the algorithm and then use the patterns on a single slice to produce the output. Then compare the filtered output with Alessandra’s baseline segmentation output. (for 2D to start with).
2. Proof of concept: same as above but for 3D
3. Merging software to a single framework: use C++ instead of Python for flexibility and for performance measurements
4. Training exploration: Use a whole data set as input not the single slice
5. Training exploration: Use as input data set provided by Alessandra that discriminates grey and white matter
6. Define the impact of the different samples on pattern selection
7. Estimate our compression capabilities for a full 3D sample
There are minutes attached to this event.
Show them.
