Software Internships
|
Company
|
DELTATEC |
|
Location
|
Ans |
|
Pubication Date
|
25 Feb 2025 |
Software Internship / TFE #1 – Video data path on embedded AI platform
Title : “Implementing a video data path on an embedded AI platform ”.
Target : Computer Science / Bachelor, Master, Engineer in Computer Science / Electronics
Type of work :
- Research and innovation 0/3
- State-of-the-art analysis 1/3
- Practical implementation 3/3
Referent : Christophe Seyler / Philipe Cornet
Internal competence center: Embedded / PC System
Job description :
DELTATEC is active in the world of machine vision: thanks to its experience in developing customized hardware around sensors and embedded systems, DELTATEC has identified a new niche in the field of smart cameras. In order to design and offer these systems, it is essential to master the entire chain: capture, processing, augmented reality and broadcasting.
The work involves the following elements:
- Building and deploying an operating system (embedded Linux) on the target platform (the platform could typically be an evaluation board including the Hailo-15 AI Vision Processor).
- Real-time video capture from a camera (typically connected to a MIPI interface).
- Real-time content processing (typically execution of neural networks on a HAILO inference chip).
- Real-time generation of overlays and mixing with the source stream (typically highlighting elements detected by the network).
- Encoding of augmented content (H.264/HEVC).
- Distribution of augmented content (typically over an IP interface).
The entire data-flow will be implemented using the GStreamer framework. Particular attention will be paid to mastering the notions of multithreading and buffering in order to guarantee efficient data flow management. Optimizing overall latency (from capture to delivery of augmented content) will be an important focus of the work.
Software Internship / TFE #2 – Mise à jour sécurisée OTA utilisant RAUC
Title : “Design and implementation of a secure OTA update system for ARM embedded systems using the RAUC framework”.
Target : Computer Science / Bachelor, Master, Computer Engineer
Type of work:
- Research and innovation 1/3
- State-of-the-art analysis 2/3
- Practical implementation 3/3
Referent Philipe Cornet / Christophe Seyler
Internal competence center: Embedded / PC System
Job description:
The work focuses on the development of a secure OTA (Over-The-Air) update system for embedded systems based on ARM architecture. The main objective is to use the RAUC framework to facilitate embedded system updates, while focusing on data security and error protection during update operations.
Embedded systems are ubiquitous in various fields such as automotive, industrial and intelligent devices. The ability to update these systems remotely without physical intervention is essential to guarantee long-term flexibility, maintenance and security. However, updating an embedded system can present challenges, particularly in terms of security and error management.
For this work, we propose to use the Robust Auto-Update Controller (RAUC) framework to facilitate ARM embedded system updates. RAUC is an open source tool designed for embedded Linux systems, offering advanced OTA update management features. It is specifically designed to address issues of robustness, security and error handling.
Data security is a top priority in any OTA update system. We will therefore look in detail at the encryption and authentication techniques used to guarantee the integrity of updates and prevent potential attacks. We will also explore key management methods and security protocols adapted to the constraints of ARM embedded systems.
With regard to fallback protection, we are asked to identify mechanisms for restoring the system to a functional state in the event of update failure or corruption. We will study memory partitioning and backup techniques, as well as integrity checking algorithms to detect errors and restore the previous state in the event of a problem.
The work will culminate in the implementation of a prototype on a Raspberry-type SBC platform.
Software Internship / TFE #3 – Accelerated Multimedia Encoding & Streaming
Title: « Development of a prototype for encoding a multimedia stream and transmitting it using standard IP protocols»
Target : Computer Science / Bachelor, Master, Computer Engineer
Type of work :
- Research and innovation 0/3
- State-of-the-art analysis 1/3
- Practical implementation 3/3
Referent: Jérôme Bayaux / Stéphane Witryk
Internal competence center: Multimédia Streaming / Compression
Job description:
In the context of remote (TV) production, IP protocols for transmitting LIVE content to and from a dematerialized infrastructure have emerged, with increasing adoption in the broadcast industry. These protocols enable not only the secure distribution of content, but also the remote processing of this content in public or private clouds.
The project involves building a prototype that implements all the software components needed to encode and transmit multimedia signals using IP protocols.
The steps in the chain include
- Encoding the video essence of the multimedia stream to be transmitted, using hardware acceleration technologies (H.264 on GPU: NVENC/NVDEC, etc.).
- Encoding the audio essence of the multimedia stream to be transmitted.
- Multiplexing of the 2 essences.
- IP packetization (the coveted protocols being RTP/RTSP/SRT/WebRTC).
- Transmission of the stream over IP.
Particular attention will be paid to mastering data movement (CPU/GPU), as well as multithreading notions to ensure efficient and effective data flow management. Optimizing overall latency (from encoding and transmission of the stream to reception of the content on the remote device) will be an important focus of the work.
Software Internship / TFE #4 – Multimedia Stream Receiver and Decoder App for Android/iOS
Title: “Development of an application for receiving and decoding multimedia streams on a tablet or smartphone (Android/iOS)”.
Target : Computer Science / Bachelor, Master, Computer Engineer
Type of work :
- Research and innovation 1/3
- State-of-the-art analysis 1/3
- Practical implementation 3/3
Referent: Yohann Vanfrachem
Internal competence center : Multimédia Streaming / Web
Job description:
The aim of this internship is to develop a mobile application running on a tablet or smartphone (Android or iOS, priority to Android) capable of receiving a multimedia stream via a standard protocol over IP (via the device's Wifi), decoding the content (H.264 or HEVC) and displaying the result in real time in the application.
Project details :
- Multimedia stream reception: Use of standard protocols over IP to receive the stream (typically RTSP, RTP, SRT).
- Content decoding: The video stream, encoded in H.264 or HEVC, will be decoded using a framework adapted to the mobile platform. This includes researching possible frameworks and selecting one of them (MediaCodec, ExoPlayer, FFmpeg, AVFoundation, VideoToolbox...).
- Result display: development of a graphical interface using a hybrid approach: the application will embed a web page for display of the decoded stream and user interaction. This includes researching possible frameworks and selecting one of them (Flutter, native React...).
- Development of the web page using modern web technologies: HTML, CSS, JavaScript, possibly using front-end frameworks such as React, Vue.js or Angular.
- Release of the developed application on the mobile platform's official store.
Software Internship / TFE #5 – Multi screen rendering with Unreal Engine
Title : “Study and Application of the Unreal Engine in a distributed rendering context”.
Target : Bachelor's degree in Computer Science / Master's degree in Industrial Computer Engineering
Type of work:
- Research and innovation 1/3
- Analysis of the state of the art 2/3
- Practical implementation 3/3
Referent : Gérald Thielen
Internal competence center: Computer Graphics
Job description :
DELTACAST markets a number of products for graphic synthesis in broadcast solutions: Augmented Reality, Virtual Reality, TV broadcast wrapping, etc. To this end, DELTACAST has developed its own distributed wrapping product capable of addressing a large number of output/display devices.
At the same time, virtual studio designers and motion designers have begun to massively adopt the Unreal graphics engine, originally developed for video games by Epic.
The work will focus on the nDisplay technology integrated into the Unreal rendering engine, enabling multiple instances of the Unreal engine distributed on several remote computers to be driven in perfect synchronicity.
In particular, the work will consist of :
- A detailed study and analysis of the concepts implemented: clock synchronization, hardware synchronization, “catch-up” mechanism, possible constraints on hardware, software and rendering scenes, etc.
- Implementation of an nDisplay-compatible infrastructure with appropriate hardware and a demonstration of the possibilities, with one or more representative scenes, including verification of the solution's correct operation in a disturbed and constrained environment (stress tests).
- Particular attention will be paid to frame-accuracy, and to the synchronization of reproduction on physical screens.
- Production of a demonstrator implementing the rendering of a scene synthesized per piece by several unreal renders and displayed per piece on several screens. The scene will include, among other things, a clock, the animation of an element crossing the scene, and possibly the playback of an offline video.
Software Internship / TFE #6 - Docker swarm / Kubernetes / Nomad
Title : “Study and comparison of container orchestration tools”.
Target : Bachelor's degree in Computer Science / Master's degree in Industrial Computer Engineering
Type of work:
- Research and innovation 0/3
- State-of-the-art analysis 2/3
- Practical implementation 3/3
Referent : Benoît Willems
Internal competence center: IT Application
Job description :
The use and deployment of IT services through Docker containers is now so widespread that Docker has become an indispensable part of an IT infrastructure.
DELTATEC, as a developer of ultra-high-fidelity electronic and IT solutions, is no exception to the rule.
DELTATEC not only has its own IT department, but also an in-house tools development department; these 2 departments work closely together to provide the company's engineers with a modern, reliable infrastructure to support all engineering work, which is spread across 3 Business Units and as many laboratories.
DELTATEC wishes to extend its internal service level, currently based mainly on a redundant cluster of virtual machines, by offering a similar service based on Docker containers.
The expected work will consist of :
- An analysis and comparison of Docker container orchestration tools (e.g. Kubernetes, Swarm and Nomad) on a number of aspects: implementation, practicality, flexibility, maintenance, consumption, reliability, etc.
- An analysis and understanding of the virtualized and containerized services running at DELTATEC.
- Drawing up a coherent proposal for implementing a new, reliable container hosting service, and migrating services from the existing infrastructure to this new service.
- The realization of one (or more if necessary) laboratory proof-of-concept.
Software Internship / TFE #7 - Enhance multimedia tooling
Title : “Enhance multimedia content analysis and testing tools”.
Target : Computer Science / Bachelor, Master, Computer Engineer
Type of work:
- Research and innovation 0/3
- State-of-the-art analysis 1/3
- Practical implementation 3/3
Referent : Jérôme Bayaux / Stéphane Witryk
Internal competence center: Compression
Job description :
In the context of our graphic design products, there is a lot of multimedia content. In particular, video decoding and rendering are ubiquitous tasks that can prove complex given the large number of video formats to be supported.
The aim of this project is first and foremost to improve our analysis capabilities, so that we can better understand the results obtained when decoding and rendering a given video.
The work consists in building an application with a graphical user interface (GUI). It will rely on independent libraries/tools (e.g. libavcodec, ffmpeg, ffprobe...) to analyze the video input.
The application will be able to perform the following actions on the videos received:
- Display “generic” metrics (resolution, colorimetry, codec, packing, transfer functions, disk speed required for playback, etc.).
- Display of metrics specific to Deltacast: for example, detecting whether video decoding or pixel conversion on the GPU is supported or not. Or analyze presentation timestamps (PTS) to check whether they are continuous or not, for example.
- Generate an analysis report in a text format that can be used in Gitlab issues.
- Secondary objective to be investigated: simple remuxing to realign video/audio essences or even PTS.
Time permitting, the generation (and automation) of tests for the internal video decoding module could also be included as an additional objective of the work.