Tricky images? Seek advice.
Imaging technology is progressing rapidly and it is becoming increasingly challenging for non-experts to stay up-to-date with the latest developments in image analysis.
To provide EPFL’s imaging community with additional support in image analysis, the EPFL Center for Imaging is running a series of bi-monthly Image Analysis Breakfasts. These data-centered meetings provide participants with a precious chance to discuss all kinds of image-analysis questions with a team of volunteer EPFL imaging experts.
The philosophy of the Image Analysis Breakfasts is not to solve imaging problems “on the spot”, although this has happened in the past! Rather, the goal is to provide participants with useful advice on the best ways to tackle their imaging problem, and on the amount of effort and resources required for doing so.
NB: The coffees and croissants that traditionally enhance our Image Analysis Breakfasts are on a short break until the sanitary situation allows for in-person meetings on campus.
During the 1-hour session, three participants present their images and their analysis problems in front of a panel of local imaging experts.
Each participant is asked to present his/her image-analysis question in 5 minutes. Then, the experts discuss the problem with the participant for 15 minutes, trying to provide tips, ideas, pointers to relevant references and/or links to similar solved problems.
The suggested resources and actions are collected in a GoogleDoc made accessible to the participant at the end of the meeting.
1. Prepare a concise 5-minutes presentation that describes your problem using our Google Slide Template.
We kindly ask you to
- avoid using domain-specific jargon;
- select typical examples that clearly illustrate the type of information you’re trying to “extract” from your images;
- make sure to have enough variability in the displayed images.
2. Send the link to your slides and to your data (typically hosted on your EPFL Google Drive) to [email protected].
We’ll then quickly get back to you to schedule a brief “preparatory Zoom call” with one of our imaging experts, who will quickly assess the suitability of your problem for the Image Analysis Breakfasts and/or may ask you to include additional information/clarification prior to your presentation at the next Image Analysis Breakfasts meeting.
3. Should the “preparatory Zoom call” be conclusive, we’ll register you to the next Image Analysis Breakfast.
Any question? See our FAQ below or contact [email protected]
A big shootout to all our enthusiastic imaging experts! This team of volunteer EPFL researchers come from diverse imaging backgrounds and have various image-processing/analysis expertise. We aim to cover most imaging modalities (from nano to macro) and software tools.
From the EPFL Center for Imaging: Daniel Sage, Olivia Mariani.
From the BIOP Platform: Olivier Burri, Romain Guiet, Nicolas Chiaruttini, Arne Seitz.
From various EPFL imaging labs: Arianne Bercowsky (SV), Pol Del Aguila Pla (STI), Sepand Kashani (IC), Matthieu Simeoni (IC), Jean-Philippe Thiran (STI), Martin Weigert (SV).
Please write us if you would like to join us!
FREQUENTLY ASKED QUESTIONS (FAQ)
No. The Image Analysis Breakfasts are currently open only to EPFL research staff (PhD Students, postdocs, scientific collaborators). EPFL Master students can attend under the supervision of their assistant, provided that free spots are available.
Yes, there is absolutely no such a thing as a stupid question/problem in image-analysis. Everybody is there to learn.
Yes, that is exactly the purpose of the breakfasts discussions: to determine the kind of effort you need to invest to solve your image-analysis problem, and to point you towards possible solutions.
Past Meetings (2020 & 2021)
Meeting 17 February 2020
- Deep learning methods for retinal cells images classification depending on disease – Timothée Laforest [STI]
- Detection of crack in walls – Amir Rezaie [ENAC]
- Bacteria classification – Kunal Sharma [SV]
- Detection of the cracks in a 3D volume of X-ray tomography – Eleni Stavropoulou [ENAC]
Meeting 16 March 2020
- Cancelled due to the Coronavirus
Meeting 20 April 2020
- Surface roughness from microCT – David Delavaux, Guillaume Anciaux, Pascal Turberg [Pixe, ENAC]
- Detection of cracks in a sequence of images– Amir Rezaie [EESD, ENAC, Prof. Beyer]
- Cells movement – Quentin Talon (LPMV, Giovanni Dietler)
- Analysis of hand written gestures – Thibault Asselborn [CHILI, IC, Prof. Dillenbourg]
Meeting 18 May 2020
- Semi-automatic tools for cracks detection – Amir Rezaie [ENAC]
- Image processing of calcium imaging – Zhang Chuanqiang [SV]
- Spherical Harmonics – Mahmoud Shaqfa [ENAC]
Meeting 15 June 2020
- Structural noise in 2-photons microscopy images – Carlos Vivar Rios and David Litvin [CHUV]
- Detection of spatio-temporal event and constriction site – Dora Mahecic [LEB, SB]
- Similarity measurement of skeletonized cracks – Amir Rezaie [EESD, ENAC, Prof. Beyer]
Monday 14 September 2020
- Lineage of E.coli bacteria – Frédéric Normandeau, McKinney Lab [SV]
- TEM image analysis of porous film – Michał Macha, A. Radenovic Lab [LBEN, STI]
Monday 16 November 2020
Computer-aided segmentation and re-association of fragmented human bones – Varlet Vincent, Lise Malfroy Camine, Negahnaz Moghaddam, Lorenzo Campana, Swiss Human Institute of Forensic Taphonomy (SHIFT) and Centre Universitaire Romand de Médecine Légale (CURML)
Microbead agglutination assay inside a microcapillary channel – Bastien Schyrr [FIMAP, IMX]
Build the lineage of e-coli bacteria – Frédéric Normandeau, McKinney Lab [SV]
Atomic islands segmentation – Karla Banjac [LSEN, SB]
Monday 18 January 2021
- Distribution of cells (Mathematica pipeline), Eva Kurisinkal, Programmable Biomaterials Lab, MX
- Fragments of human bones, Lise Malfroy, CHUV, UNIGE
Monday 22 March 2021
- Semantic 3D segmentation of low-contrast, artefact-corrupted, TEM tomographic reconstructions, Robin Girod, INE, MX
- quantify the vascularization of a drosophila tissue (coverage, branching…), Samuel Rommelaere, SV