Acquire direct experience at the intersection of machine‑learning systems and cryptographic research by contributing to the development of methodologies tailored to the secure and efficient processing of audio and visual media using advanced primitives such as fully homomorphic encryption (FHE), zero‑knowledge proofs (ZKPs), and secure multiparty computations (MPC). #machine-learning #fully-homomorphic-encryption #zero-knowledge-proof #cryptography

Description
As trust is increasingly a critical concern in distributed systems, cryptographic techniques like Fully Homomorphic Encryption (FHE), Zero-Knowledge Proofs (ZKP), and Multi-Party Computation (MPC) provide mechanisms to ensure trustworthy computation. However, their computational cost typically scales with the volume of data processed, which poses major challenges when processing media content like audio or video.

In this project, you will investigate methods to operate directly on compressed media formats within cryptographic frameworks. You will explore the design of novel media formats that are amenable to cryptographic deployment, using machine‑learning techniques to automatically find optimized media representations.

This project is at the intersection of research in Machine Learning, media codecs, and applied cryptography. We especially welcome applicants with a background in ML. Ideally, this internship can become an application case of your research and can become an integral part of your PhD, as well as result in a paper.

Your responsibilities:

Acquire comprehensive familiarity with diverse audio and visual media codecs.
Conduct a critical evaluation of the computational overhead incurred when processing these formats using cryptographic protocols.
Employ machine‑learning–driven approaches to define novel media representations possessing characteristics conducive to secure cryptographic processing.
Consolidate your research in a scientific publication.

Your skills and experience:

Student enrolled in Ph.D. Computer Science/Engineering.

Proficient experience with contemporary machine‑learning methodologies and training pipelines.

A basic knowledge of FHE, ZKP, or MPC is advantageous.

A basic knowledge of media codecs is a plus.

Good programming skills in Rust, C, or C++ and in Python or Golang.

Language skills: English

You must relocate to Belgium for the duration of the internship.

This is a paid internship.