FALCON · FAst and energy efficient Learned image and video CompresiON
„Хоризонт 2020“ — Действия „Мария Склодовска-Кюри“
- Период
- 2022-03-01 → 2024-02-29
- Финансиране от ЕС
- 190 681 €
- Участници
- 1
- Схема
- MSCA-IF
Линиите свързват координатора с партньорите.
Накратко на български
Методите за компресиране на изображения и видео чрез машинно обучение се изследват, за да станат по-бързи и качествени. Това помага за намаляване на енергийния разход и вредните емисии, тъй като видеотрафикът заема голяма част от интернет обмена.
Кратко обяснение, генерирано от езиков модел по текста на CORDIS. Оригиналът е по-долу.
Резултати накратко
FAst and energy efficient Learned image and video CompresiON
The ever-increasing demand for image and video-based applications motivates innovations to improve multimedia compression performance, and support modern multimedia systems. The emerging solutions exploit the power of Machine Learning to enable higher image/video compression rates, and higher quality for such systems. However, these solutions come with a high computational complexity, which translates to higher energy consumption, difficulty of deployment in services and consumer devices, and higher carbon footprint. Multimedia services are among the most demanding applications. Video-based traffic constitutes around 80% of the Internet traffic, and is responsible for around 1% of the global greenhouse gas emission. Hence, it is essential to develop solutions to enhance the compression efficiency of compression systems, while reducing the computational complexity to comply with important energy consumption policies. To this end, this project, FALCON, studies novel solutions for fast and energy efficient learning-based compression, that support modern multimedia systems. The overall objective is to improve the compression and quality, while reducing the complexity for faster solutions. The objectives of the project are achieved through the following main directions: (1) Designing fast and efficient methods for multimedia and compression systems. This lowers the complexity, while trying to keep the compress/quality. (2) Designing advanced learning-based methods for multimedia and compression systems. This enable higher compression efficiency, with similar complexities. (3) Optimizing compression systems based on human psycho-vision. This removes unnecessary information or operations that cannot be distinguished by human observers.
Текст от CORDIS, на английски · Данни: CORDIS, © Европейски съюз
Цел на проекта
The emerging Learned Compression (LC) methods show great potential to revolutionize image/video compression, and major media industries are investing heavily in this field. However, the high computational complexity of these methods makes it difficult to employ them in consumer devices, and this obstacle discourages using them in future compression standards, such as JPEG and MPEG, despite their superior performance compared to traditional methods. This project will investigate a novel framework for developing fast and energy-efficient Deep Learning-based compression. We will develop methods that (1) greatly improve the compression efficiency of LC, and (2) significantly reduce its computational complexity and energy consumption. Given the huge share of video industry in global Greenhouse gas emission, this will be a big step towards important EU policies such as the Paris agreement and the EU Green Deal. The objectives of the project are achieved via: (i) splitting the coding into smaller tasks, (ii) investigating efficient learning methods (including Operational Neural Networks, an invention of the supervisor of the project), and (iii) integrating human perception into image/video coding.The experienced researcher holds a PhD in computer engineering, during which he worked on accelerating the encoding process of compression standards. He has a background and skill-set in hardware engineering, signal processing, media technology, and machine learning, which is necessary for this interdisciplinary project. The project will be carried out under the supervision of an internationally famous scientist who has extensive experience in both machine learning and video compression. The host institution in Finland has a long experience in EU funding and collaborations with industries. The results and findings will be published in top international journals and conferences. Moreover, some findings will be considered for possible exploitation in future MPEG/JPEG standards.
Оригинален текст от CORDIS (на английски).
Участници
- TAMPEREEN KORKEAKOULUSAATIO SR · TampereКоординаторФинландия
Връзки
Данни: CORDIS, © Европейски съюз