Master AI Internship – Advanced Image Segmentation & Shape Completion

Master AI Internship – Advanced Image Segmentation & Shape Completion

ABOUT THE JOB AND ABOUT THE TEAM

In this project assignment, you will work on an advanced image segmentation challenge that sits at the intersection of design intuition and algorithmic precision. The main objective is finding ‘natural’ ways to close open shapes in designs, in an almost pixel-perfect way—closely mimicking how Vlisco designers interpret, complete, and finalize these shapes.

You will collaborate closely with a multidisciplinary team of experts in data science, technology, and design, to translate human design logic into scalable technical solutions. Vlisco provides a balanced dataset containing ground truth designs, including “plukkanalen” and blockings, allowing you to focus on developing and validating innovative segmentation methods within a real production environment. The workplace is located in the city center of Helmond, within walking distance of the train station.

KEY RESPONSIBILITIES

• Develop methods to close gaps in semi-open shapes in a natural way, using techniques such as Bezier curve-based interpolation or similar curve-fitting strategies;
• Investigate how designers intuitively complete shapes and try to replicate this algorithmically;
• Consider approaches like Fourier-based techniques for patterns that require frequency-domain analysis, such as hatched patterns;
• Ensure solutions are robust across diverse designs;
• Explore whether curve completion should occur during preprocessing or post-segmentation (e.g., after SAM) and evaluate cost (post-segmentation should be fast) vs. benefits (but is expected to be more accurate);
• Evaluate segmentation results using a balanced dataset with ground truth blockings;
• Compare outcomes with reference blockings and validate results in collaboration with designers;
• Integrate the developed solution into a prototype plugin;
• Where possible, implement the solution as a standalone feature within the Vlisco Photoshop plugin (e.g., as a preprocessed layer).

WHAT WE ARE LOOKING FOR

• A student in Computer Science, Artificial Intelligence, Data Science, Computer Vision, or a related technical field;
• Strong interest in image segmentation, shape analysis, and bridging human intuition with algorithmic solutions;
• Experience or interest in curve modeling, geometric algorithms, or signal processing techniques;
• Good knowledge of English (spoken and written);
• Familiarity with Python and image processing or machine learning libraries is a plus;
• Analytical, creative, and detail-oriented mindset;
• Comfortable working in a multidisciplinary environment with designers and engineers;
• Available for 32-40 hours for a period of 6 to 9 months.

IN RETURN WE OFFER YOU:

• Internship compensation;
• A workplace located in the city center of Helmond, within walking distance of the train station;
• An exceptional learning opportunity;
• Collaboration and stakeholder management.