Student Projects

DNA can be used for storage of information the same way the genetic codes of most living entities, including humans, are stored in their DNA. There are several advantages behind such an approach, such as a much higher storage density, long term preservation capability and better energy efficiency. The underlying information in DNA is represented in a quaternary code (AGCT) instad of a binary code (01). This calls for completely new approaches to efficiently code information in a DNA compatible manner. 

The goal of this project is to study alternatives approaches proposed in the state of the art to store informaiton in DNA and to come up with an end-to-end image compression simulator by taking advantage of publicly accessible implementations.  

The following tasks should be performed during the project:

• Study the relevant state of the art relevant in DNA storage and coding.
• Identify existing source code for DNA storage and analyse them.
• Design and implement a simulator of image compression for DNA storage based on state of the art implementations  
• Analyse the performance of the simulator.

Requirements: Basic knowledge of signal and image processing. Good programming skills.

Contact: Touradj Ebrahimi

Group: Prof. Touradj Ebrahimi

Suitable for: Bachelor Semester Project, Master Semester Project or Master Thesis in Electrical Engineering, Communication Systems, Computer Science, Digital Humanities, Mechanical Engineering, Micro Engineering, Management of Technology and/or equivalent.

Number of students: One

Due to the increasing spread of doctored or synthetic contents on the Internet and their impact on the dissemination of fake news over social networks, detecting manipulated content has become a major challenge in both academic and professional communities. Major companies have joined forces to organize challenges with the goal of helping in the process of creating widely accessible tools and solutions to detect malicious modifications of multimedia contents.

One of the most important and recent actions was the Deepfake Detection Challenge organized by Facebook and Microsoft, with the involvement of many academic research groups. The organizers hoped that this challenge would result in new technologies for detecting AI-generated videos which can later be used on social networking platforms and/or by journalists. This illustrates the major concerns of large companies about the danger of AI-assisted content manipulations. 

In this project, we tackle the deepfake detection problem by training several convolutional neural networks (CNNs) in a supervised fashion. Finally, the ensembling of different trained CNNs will be studied.

In particular, two main objectives will be pursued in this project. The first aims at finding existing and publicly available deepfake datasets. The second aims at training deep neural networks using the above datasets for the task of deepfake detection. The following tasks should be performed by the student:

• Review the state of the art deepfake detection methods 
• Study the state of the art deepfake creation methods and find/generate their corresponding dataset which can further be used for training of CNNs.
• Run/Adapt/Create a program to detect deepfake images and videos
• Investigate the most common performance metrics
• Assess the performance of the trained models against several datasets
• Document the code and write a report on the project

Requirements: Good skills in Python programming. Background in deep learning and image processing.

Contact: Yuhang Lu

Group: Prof. Touradj Ebrahimi

Suitable for: Master Semester Project or Master Thesis in Electrical Engineering, Communication Systems, Computer Science, or equivalent.

Number of students: One

The recent advances in deepfake technology as well as their potential to spread misinformation have raised an alert on the research community. For this reason, several algorithms for automatic detection of deepfakes have been developed, particularly targeting examples where humans are not able to discern between fake and real. Among the different studied multimedia modalities, audio deepfakes have received special attention since they allow the malicious impersonation of any personality through phone lines, a technique that has already been employed for attempts of fraudulent bank transfers and other attacks.

The goal of this project is to study state-of-the-art solutions for the generation and detection of audio deepfakes. The student will explore the research literature and select different methods to produce a comprehensive audio deepfake dataset. The produced dataset will then be examined by different detection methods in order to assess their performance.

The following tasks should be performed during the project:

• Study the state-of-the-art in audio deepfake generation and detection.
• Select algorithms with available source code to produce the dataset.
• Produce a comprehensive dataset with fake and real audio samples of human speech.
• Run different detection methods on the produced dataset.
• Evaluate and compare the performance of the detection methods, deriving conclusions regarding their strengths and weaknesses.
• Document all the development process and source code.

Requirements: Background on deep learning. Good skills in programming.

Contact: Davi Lazzarotto, Yuhang Lu

Group: Prof. Touradj Ebrahimi

Suitable for: Bachelor Semester Project, Master Semester Project or Master Thesis in Electrical Engineering, Communication Systems, Computer Science, or equivalent.

Number of students: One.

In recent years, the deepfake technology has achieved remarkable progress in various domains and different data modalities, including but not limited to audio, image, video, etc. For example, one can create convincingly mimic voices to deceive listeners through telephone. The deep learning tools and open-source software have simplified the creation of such forgery content. Therefore, the malicious potential of audio deepfakes highlights the urgent need for effective detection methods.

The primary goal of this project is to explore and implement state-of-the-art deep learning techniques for the detection of audio deepfakes. The student will conduct an in-depth investigation of existing methods and datasets in this area, develop a detection system using Python, train the detector on diverse datasets, and rigorously evaluate of systems’ performance in different test scenarios.

The following tasks should be performed during the project:

• Investigate the state-of-the-art methods and datasets for audio deepfake detection
• Implement one or more deep learning-based audio deepfake detection method
• Train the detection system with appropriate datasets
• Evaluate the detector’s performance and compare its effectiveness on various datasets
• Document the code and results and write a report on the project

Requirements: Background on deep learning. Good skills in programming. 

Contact: Yuhang Lu, Davi Lazzarotto

Group: Prof. Touradj Ebrahimi

Suitable for: Bachelor Semester Project, Master Semester Project or Master Thesis in Electrical Engineering, Communication Systems, Computer Science, or equivalent.

Number of students: One.

3D media content can be represented in a variety of formats. Gaussian splatting has been recently proposed as a method to generate 3D representations from a set of images depicting an object or scene from different angles. When compared to neural radiance fields (NeRFs), gaussian splatting boasts lower training and inference times at similar levels of quality, and thus have been quickly adopted by the research community.

The goal of this project is to study state-of-the-art solutions for the generation of gaussian splatting models. The student will explore the research literature and select different methods to produce a comprehensive dataset from sets of pictures depicting different objects and scenes. These images can be either produced by the student or extracted from the internet. The produced dataset will then be carefully examined to derive conclusions about the strengths and weaknesses of different methods, as well as about best practices for the generation of high-quality gaussian splatting models.

The following tasks should be performed during the project:

• Study the state-of-the-art in 3D gaussian splatting.
• Select algorithms with available source code to produce the dataset.
• Generate sets of pictures of different objects to produce the dataset, either acquired by the student or extracted from the internet.
• Generate gaussian splatting models with the selected methods.
• Methodically compare the subjective quality of the models, exploring differences between different methods as well as defining best practices for the generation of the models.
• Document all the development process and source code.

Requirements: Background on image processing and machine learning. Good skills in programming. 

Contact: Davi Lazzarotto

Group: Prof. Touradj Ebrahimi

Suitable for: Bachelor Semester Project, Master Semester Project or Master Thesis in Electrical Engineering, Communication Systems, Computer Science, or equivalent.

Number of students: One.

3D media content can be represented in a variety of formats. Point clouds depict objects as sets of unconnected points and have been widely studied as a 3D imaging modality. However, point clouds can contain a vast amount of data, and thus effective compression solutions are needed. Deep learning has been widely explored for this task, notably with the use of sparse convolutional networks, but recent developments on the field still leave room for performance improvement.

The goal of this project is to attempt to enhance the performance of current learning-based models for point cloud compression. The student will get acquainted with recent advances on the fields of generative AI, specially for 3D content, and develop ways to translate them to state-of-the-art algorithms for point cloud compression in conjunction with their supervisor.

The following tasks should be performed during the project:

• Study the state-of-the-art in point cloud generative AI and learning-based point cloud compression.
• Design and implement algorithms for point cloud compression based on the research on the state of the art.
• Assess the rate-distortion performance of the proposed methods.
• Compare the performance of the implemented approaches to state-of-the-art solutions both in terms of objective and subjective quality.
• Document all the development process and source code.

Requirements: Background on image processing and machine learning. Good skills in programming. 

Contact: Davi Lazzarotto

Group: Prof. Touradj Ebrahimi

Suitable for: Bachelor Semester Project, Master Semester Project or Master Thesis in Electrical Engineering, Communication Systems, Computer Science, or equivalent.

Number of students: One.

With recent advancements in generative AI, models that generate high-quality images are becoming increasingly important for applications such as advertising, entertainment, and content creation. However, these models often require fine-tuning on specific datasets to improve their performance and relevance to the task at hand. The goal of this project is to develop a comprehensive dataset of images that will be used to fine-tune visual generative AI models, improving their ability to produce more contextually relevant and visually appealing content.

The main objective of this project is to create and organize a dataset that enhances the performance of generative AI in producing images. This involves studying the current state-of-the-art in visual generative models, selecting or capturing relevant images, and structuring them in a way that supports fine-tuning processes.

The following tasks should be performed during the project:

• Study the state-of-the-art in visual generative AI and dataset requirements.
• Create a dataset of images suited for fine-tuning generative models, ensuring diversity in image types, resolutions, and categories.
• Implement pre-processing techniques to standardize and annotate the dataset for machine learning tasks.
• Develop and document scripts for data collection, cleaning, and augmentation to ensure the dataset is ready for model training.
• (Optional) Fine-tune existing generative models using the newly created dataset, and evaluate their performance.
• Document the entire development process, including data curation steps, pre-processing techniques, and model fine-tuning, along with the source code.

Requirements: Background on image processing and machine learning. Good skills in programming. 

Contact: Dr. Evgeniy Upenik, Yuhang Lu, Davi Lazzarotto

Group: Prof. Touradj Ebrahimi

Suitable for: Bachelor Semester Project, Master Semester Project or Master Thesis in Electrical Engineering, Communication Systems, Computer Science, or equivalent.

Number of students: One.

Recent advancements in generative AI have significantly improved the ability to create visual content, but ensuring that the generated images accurately convey the intended meaning remains a challenge. The semantics of AI-generated visual content must align with user intentions, making it crucial to develop reliable methods for assessing this alignment. The goal of this project is to develop and evaluate methodologies for assessing the semantic relevance of AI-generated images.

The objective of this project is to create and validate methodologies for assessing the semantic relevance of images generated by AI models. This involves designing evaluation frameworks, gathering feedback through subjective (human) assessments, and exploring objective metrics to quantify how well the images match the intended semantic content.

The following tasks should be performed during the project:

• Study the state-of-the-art in generative AI models for images, with a focus on existing methods for evaluating semantic relevance.
• Develop subjective assessment methodologies by designing experiments where human evaluators judge the semantic accuracy of AI-generated images.
• Explore objective metrics that can quantitatively measure semantic relevance, such as image-text alignment or other machine-learning-based approaches.
• Implement subjective (and optionally objective) evaluation methods and apply them to assess the performance of existing AI models.
• Evaluate and compare the performance the developed methodologies, identifying their strengths and weaknesses.
• Document all methodologies, results, and source code, providing a clear guide for future research in this area.

Requirements: Good background on image processing and machine learning. Good skills in programming. 

Contact: Dr. Evgeniy Upenik, Yuhang Lu, Davi Lazzarotto

Group: Prof. Touradj Ebrahimi

Suitable for: Bachelor Semester Project, Master Semester Project or Master Thesis in Electrical Engineering, Communication Systems, Computer Science, or equivalent.

Number of students: One.

As generative AI models become more advanced, the quality of images they produce has seen significant improvements. However, while these models excel at generating technically correct images, they often lack the ability to consistently create visually appealing content that resonates with human users. Visual appeal, a critical factor in industries such as advertising, entertainment, and art, requires fine-tuning of generative models. The goal of this project is to develop and implement methods to fine-tune generative AI models, enhancing their ability to create visually appealing images that align with human aesthetic preferences.

The main objective of this project is to improve the visual appeal of images generated by AI. This will involve developing fine-tuning methods, using objective metrics and human feedback to adjust the AI models, and evaluating their performance based on perceptual visual quality.

The following tasks should be performed during the project:

• Study the state-of-the-art in generative AI models for image creation, focusing on the techniques used to enhance visual appeal.
• Design and implement fine-tuning techniques for existing generative models, incorporating visual quality metrics (and optionally user feedback).
• Fine-tune a selected generative AI model using the developed methods and evaluate the improvement in visual appeal.
• Compare the performance of the fine-tuned model against the original model based on the developed visual appeal metrics.
• Document the entire development process, including fine-tuning methods, evaluation procedures, and source code.

Requirements:Good background on image processing and machine learning. Good skills in programming.

Contact: Dr. Evgeniy Upenik, Yuhang Lu, Davi Lazzarotto

Group: Prof. Touradj Ebrahimi

Suitable for: Bachelor Semester Project, Master Semester Project or Master Thesis in Electrical Engineering, Communication Systems, Computer Science, or equivalent.

Number of students: One.

Recently, public interest in privacy protection has increased dramatically. However, there is a general believe that protection of privacy will restrict online benefits of users. Therefore, protection of privacy in such a way that does not distract online habits of people is needed. This project focuses on visual privacy protection in images. Specifically, the intention of the project is to develop a mobile (iOS platform) application that would be able to obfuscate personal visual information in an image in a secure and recoverable way and share images via online social networks in a secure way.

The following tasks should be performed during the project:

• Research and review the existing visual privacy protection tools, as well as the way to share and manage secure content in social networks.
• Design an app on smartphone with iOS operating system. An iPhone will be provided by the lab.
• Minimal requirements of the app include:
  • Implementation of security processing (e.g. scrambling) for images on the mobile side.
  • Multi-region processing on image using touch screen.
• Implementation of a simple key management system.

Requirements: Basic knowledge of image processing, good programming skills in Objective-C, experience in iOS development.

Contact: Davi Lazzarotto

Group: Prof. Touradj Ebrahimi

Suitable for: Bachelor Project or Master Semester Project in Electrical Engineering, Communication Systems, or Computer Science.

Number of students: One.

Recently, public interest in privacy protection has increased dramatically. However, there is a general believe that protection of privacy will restrict online benefits of users. Therefore, protection of privacy in such a way that does not distract online habits of people is needed. This project focuses on visual privacy protection in images. Specifically, the intention of the project is to develop a mobile (Android platform) application that would be able to obfuscate personal visual information in an image in a secure and recoverable way and share images via online social networks in a secure way.

The following tasks should be performed during the project:

• Research and review the existing visual privacy protection tools, as well as the way to share and manage secure content in social networks.
• Design an app on smartphone with Android operating system. An Android phone will be provided by the lab.
• Minimal requirements of the app include:
  • Implementation of security processing (e.g. scrambling) for images on the mobile side.
  • Multi-region processing on image using touch screen.
• Implementation of a simple key management system.

Requirements: Basic knowledge of image processing, good programming skills in Java, experience in Android development.

Contact: Davi Lazzarotto

Group: Prof. Touradj Ebrahimi

Suitable for: Bachelor Project or Master Semester Project in Electrical Engineering, Communication Systems, or Computer Science.

Number of students: One.

The rapid growth of photo sharing through social media raises serious questions related to ownership, privacy and access to shared images. From the user perspective, effective privacy protection tends to impose restrictions on how users share and access pictures, making privacy protection unattractive. To address such issues, MMSPG has developed ProShare, a mobile App through which pictures can be protected, shared and made selectively accessible in a transparent manner while incurring minimal distraction to the user. To date a large effort has been invested in the development and implementation of the ProShare mobile App. Less attention has been directed at the server side realisation of ProShare.

The objective of this student project is to enhance the server side implementation of the ProShare service.

The following tasks should be performed during the project:

• Study and understand the ProShare service and its implementation (both server side and client side)
• Review server side architecture and compare this to state-of-the-art implementations for similar services
• Propose modifications and enhancements to the existing server side implementation
• Implement a migration process allowing to move the ProShare server to a new computing platform
• Implement robust and reliable session management
• Propose and implement additional service features
• Implement back-end tools for the analysis of usage and user statistics

Requirements: Good communications skills. Good understanding of web server technologies including Apache, MySQL and PHP. Good abilities to think at the systems level.

Contact: Davi Lazzarotto

Group: Prof. Touradj Ebrahimi

Suitable for: Bachelor Semester Project in Computer Science, Communications Systems, Electrical Engineering, or equivalent.

Number of students: One