Master Thesis - Approximative Computing for CNNs

Location: Lund, Sweden; on-site. The thesis is open only to students affiliated with a Swedish university or college, directly or through an exchange program.

Company

hirify.global develops network technology and intelligent solutions for video surveillance, access control, intercom, and audio systems.

What you will do

• Investigate the impact of approximate arithmetic circuits on CNNs for object detection and semantic segmentation.
• Evaluate how approximate multipliers affect detection and segmentation performance.
• Quantify trade-offs between network accuracy, estimated power consumption, and hardware area.
• Explore layer-wise multiplier selection, calibration, and fine-tuning methods for recovering lost accuracy.
• Use open-source libraries of approximate multipliers and simulation frameworks to turn experimental results into well-founded conclusions.

Requirements

• Master’s-level studies in electrical engineering, computer engineering, computer science, or a related field.
• Solid programming skills in Python and experience with a deep learning framework such as PyTorch or TensorFlow.
• Basic understanding of convolutional neural networks and their training and evaluation.
• Coursework or experience in digital design or computer architecture, including computer arithmetic.
• Affiliation with a Swedish university or college is required.
• Two students are expected to work together; paired applicants should identify their co-applicant.

Nice to have

• Experience with object detection or semantic segmentation models.
• Familiarity with neural network quantization or hardware-aware model optimization.
• Experience with Verilog, VHDL, or synthesis tools.
• Knowledge of C, C++, or CUDA.
• Interest in energy-efficient computing and edge AI.

Culture & Benefits

• Fixed-term thesis worker position for two Master’s students completing 30 credits each.
• Collaborative work at the intersection of hardware and machine learning.
• Applications are accepted in Swedish and English.
• Applications must include a CV and university or college grade summary.

Hiring process

• Apply through the thesis proposal advertisement.
• Submit one application per student when applying as a pair, clearly naming the co-applicant.
• Background checks may apply to certain roles; candidates will be notified before any action is taken.