Internship - Test Engineer - Design of a reticle detection module via AI - M/F
Job Description
- First experience in embedded computers.
- Pursuing a BAC+5 degree in system or software development.
- Seeking a 6-month internship.
- Master Python scripting language.
- Rigorous and naturally curious.
- Demonstrate good interpersonal skills.
- Collect use cases for reticle verification.
- Modify operational code to indicate incrustations on images.
- Validate video sequences based on use cases.
- Create a large image database for algorithm training.
- Study and select AI algorithms.
- Design the detection algorithm.
- Compare algorithm results with current solutions.
- Technological expertise.
- Attention to employee work-life balance.
- Inclusive and benevolent environment.
- Recognized societal and environmental commitment (Thales Solidarity, CAC 40 ESG index).
- Opportunity to learn alongside passionate experts.
- Quality mentorship and guidance.
- Challenging work.
- Conventioned internship with gratification based on study level.
- Thales is a Handi-Engagée company, recognizing all talents and diversity.
Your daily life
At the heart of Saint Quentin en Yvelines, the Elancourt site is one of Thales' most important technological centers of excellence. The diversity and technicality of the systems and equipment mobilize expertise in more than 80 professions in defense, high-tech security, and digital services solutions.
You will join the DCL (Computer & Software) department of the Competence Center whose main mission is to validate the functions of our airborne system computers.
It is characterized by the diversity of its professions and its multidisciplinary teams (System, Functions, Software, Sensor), as well as by the development of new complex optronic products.
By joining us, you will be entrusted with the following missions:
Problematic:
- Difficulties in automating the verification of reticles (embedded in images intended for the Rafale pilot) having variable shapes (e.g., shapes projected on surfaces) or superimposed reticles.
- Non-automated verifications are quite heavy: requiring an operator.
Step 1: Collection of use cases
- We modify the operational code so that it indicates the incrustations present on each image. An operator validates a video sequence by marking it valid/invalid according to a use case.
Each use case proposes the presence, absence, superposition of reticles and incrustations with/without background image. This creates a large image database.
The algorithm is trained. It is then used on unlabeled images in non-regression to determine if the incrustation is present or not.
Step 2: Study and choice of AI algorithms, design of the algorithm
- For example, Tesseract = Free OCR engine (by Google).
- Combined with an LLM like LLaMA, Mistral, Mixtral via API.
- Usable with frameworks like LangChain or Haystack.
Step 3: Comparison of results with those of the current solution
Your profile
Are you pursuing a BAC+5 level course in system or software development and are you looking for an internship for a duration of 6 months?
Do you master the Python scripting language?
Ideally, you have a first experience in the field of embedded computers.
Are you a rigorous and naturally curious person?
Do you demonstrate good interpersonal skills?
The team's word
An internship within our engineering department is an opportunity to learn alongside passionate individuals who are keen to share their expertise.
We believe that every intern deserves quality support: join us to be trained, guided, and challenged by our teams.
All our internships are conventional and subject to remuneration, the amount of which is determined according to your level of study.
Thales, a Handi-Engagée company, recognizes all talents. Diversity is our best asset. Apply and join us!
