Master thesis: Semantic Communications with Foundation Models for Vehicle and Road User Detection

Technology, Data & Digital · Software & Web Development · Software Engineering · Data Science

In short

This Master's thesis project investigates semantic communications for vehicle and road user detection using foundation models like CLIP and DINOv2. The research aims to compress and transmit task-relevant spatial features over noisy wireless channels, evaluating performance against various baselines and channel conditions.

Responsibilities

Build and evaluate a semantic-communication system that extracts spatial features from images using a frozen CLIP-ViT-B/16 or DINOv2-ViT-B/14 model.
Compresses them through a learned linear bottleneck, quantizes and transmits them over a simulated additive white Gaussian noise (AWGN) channel, and reconstructs bounding boxes and class labels with a lightweight decoder.
Review semantic communications, joint source-channel coding (DeepJSCC), foundation models, and JPEG compression.
Extract patch-level features from the selected COCO classes.
Implement the learned bottleneck, quantization, AWGN channel simulation, and lightweight anchor-free detection decoder.
Train the bottleneck and decoder end-to-end with noise injection at random signal-to-noise ratio (SNR) levels.
Compare both foundation-model approaches with a task-specific encoder-decoder and standard baselines such as JPEG.
Analyze results in a communications context, including mean average precision (mAP) versus SNR and per-class performance.

Requirements

Pursuing a master’s degree in Electrical Engineering, Computer Science, Telecommunications, or a related field.
Strong programming skills in Python and PyTorch.
Good understanding of deep-learning concepts, including CNNs, transformers, training procedures, and loss functions.
Ability to work independently, design experiments, and critically interpret results.
Fluency in written and spoken English.

Desired Qualifications

Familiarity with object detection (e.g., YOLO, Faster R-CNN, or FCOS), wireless communications (e.g., SNR and AWGN), or pretrained vision models (e.g., CLIP, ViT, or DINOv2) is an advantage.

Benefits

Outstanding opportunity to use your skills and imagination to push the boundaries of what’s possible.
Build solutions never seen before to some of the world’s toughest problems.
Be challenged, but you won’t be alone.
Joining a team of diverse innovators, all driven to go beyond the status quo to craft what comes next.
Encouraging a diverse and inclusive organization is core to our values at Ericsson.
We champion diversity and inclusion in everything we do.
Collaborating with people with different experiences drives innovation, which is essential for our future growth.
We encourage people from all backgrounds to apply and realize their full potential as part of our Ericsson team.
Ericsson is proud to be an Equal Opportunity Employer.

Skills

PythonPyTorchDeep LearningCNNsTransformersObject DetectionWireless CommunicationsCLIPDINOv2ViT

#Master Thesis#Semantic Communications#Foundation Models#Vehicle Detection#Road User Detection#Wireless Networks#5G-Advanced#6G#CLIP#DINOv2#Object Detection#Deep Learning#Python#PyTorch#Computer Science#Electrical Engineering#Telecommunications

About Ericsson

Pioneering technology for connecting billions of people worldwide

Our purpose \nTo create connections that make the unimaginable possible.\n\nOur vision\nA world where limitless connectivity improves lives, redefines business and pioneers a sustainable future.\n\nOur values\nPerseverance, professionalism, respect and integrity.\n\nThe future is a place for purpose & vision – ours are clear, and we invite partners, customers and consumers to join us in our journey. \n\nFor a brighter future. For all. Let's #ImaginePossible

See all open jobs at Ericsson Read more about Ericsson Website