Research & Education

dAIEDGE is a Network of Excellence dedicated to advancing Europe's innovation and technology base in AI. It focuses on distributed, trustworthy, efficient, and scalable AI at the Edge, operating under the European Lighthouse for AI. The initiative aims to foster the exchange of ideas and concepts on next-generation AI, connecting ecosystem actors to guide the European Commission and EU in future AI developments. Key objectives include driving Edge AI scientific progress through foundational research, uniting the European AI community with exchange programs, and developing a Strategic Research Agenda. dAIEDGE also showcases technological advances through use cases like warehouse optimization and satellite imagery analysis, and provides a virtual innovation infrastructure for Edge AI experimentation.

Global-Flow-Local-Attention is an open-source model designed for deep image spatial transformation, primarily focused on person image generation and animation. It leverages global flow and local attention mechanisms to achieve flexible applications such as pose-guided person image generation, pose-guided person image animation, face image animation, and view synthesis. The project provides source code, pre-trained weights, and demo scripts for quick exploration and implementation. Users can get started by installing Python, PyTorch, and CUDA dependencies, then downloading pre-trained models for various tasks like fashion image generation, video animation, and novel view synthesis. The tool is suitable for researchers and developers interested in advanced image manipulation and generation techniques.

HAT (Hybrid Attention Transformer) is an advanced tool for image super-resolution and restoration, leveraging a transformer-based architecture to significantly enhance image quality and detail. Developed by XPixelGroup, it has been featured in CVPR2023 and TPAMI, highlighting its research-backed capabilities. The tool provides various models, including those optimized for real-world super-resolution and GAN-based enhancements for sharper results. Users can test models with provided configurations, and it supports a tile mode for inference with limited GPU memory. HAT is built on PyTorch and BasicSR, offering a robust solution for researchers and developers in image processing.

libMultiRobotPlanning is a C++(14) library designed for task and path planning in multi-robot/agent systems. It provides a collection of highly templated search algorithms optimized for performance. The library includes single-robot algorithms such as A*, A* epsilon (focal search), and SIPP (Safe Interval Path Planning). For multi-robot scenarios, it supports Conflict-Based Search (CBS), Enhanced Conflict-Based Search (ECBS), and their variants with Optimal Task Assignment (CBS-TA, ECBS-TA), as well as Prioritized Planning using SIPP. Additionally, it offers assignment algorithms like minimum sum-of-cost (flow-based) and Best Next Assignment. The library is open-source, comes with useful examples, and is built for researchers and developers working on complex multi-robot coordination problems.

Nematus is an open-source neural machine translation toolkit developed by EdinburghNLP, built using Tensorflow. It provides robust support for both RNN and Transformer architectures, making it versatile for various machine translation tasks. Key features include support for advanced RNN architectures with arbitrary input features, deep models, and various dropout techniques. For Transformer architectures, it offers arbitrary input features and DropHead for attention head dropout. The toolkit also includes multi-GPU support, documentation, label smoothing, early stopping, and the ability to resume training. It provides batch decoding, n-best output, and scripts for scoring and rescoring, along with a server mode. Nematus also stores model hyperparameters, vocabulary files, and training progress in JSON format, and offers pretrained models for 13 translation directions.

pixelSplat provides the code for generating 3D Gaussian splats from image pairs, a method for scalable and generalizable 3D reconstruction. Developed by David Charatan et al. and presented at CVPR 2024, this tool allows users to reconstruct 3D scenes from 2D images. The codebase has been updated to reflect the camera-ready version of the paper, including architectural improvements like per-image self-attention in the epipolar transformer, leading to slightly better results across datasets. It supports running with an arbitrary number of views, though requiring significant GPU memory for more complex setups. The project also offers pre-trained checkpoints and scripts for dataset conversion and evaluation.

Blink Technologies, Inc. develops advanced AI-based eye tracking software and technology, aiming to create seamless interaction between human eyes and the digital world. Their non-invasive solution is compatible with any camera technology and operating system, including Windows, iOS, macOS, Android, and Linux. It can be deployed locally (edge-based) or remotely (cloud-based), offering flexibility for various applications. Blink's technology is applied across multiple sectors, from redefining online shopping experiences with gaze and fixation data to enhancing automotive safety through driver visual attention monitoring. They also provide solutions for touchless panel operation, 3D display interaction, and advanced eye tracking authentication, moving beyond traditional facial recognition for enhanced privacy and security.

RefineDet is an open-source implementation of a single-shot refinement neural network designed for object detection tasks. Published at CVPR 2018, this method aims to surpass the accuracy of traditional two-stage object detection approaches while preserving the computational efficiency characteristic of one-stage methods. The repository provides comprehensive code for training and evaluating RefineDet models on various datasets, including PASCAL VOC and MS COCO. Users can leverage pre-trained models based on VGG-16 and ResNet-101 architectures, and the system supports both single-scale and multi-scale testing strategies. It includes detailed instructions for installation, data preparation, training, and evaluation, making it a valuable resource for researchers and developers in computer vision.

SNIPER, also known as AutoFocus, is an efficient multi-scale training and inference algorithm designed for instance-level recognition tasks such as object detection and instance-level segmentation. It significantly speeds up multi-scale training by selectively processing context regions around ground-truth objects, called 'chips', operating on low-resolution data. This memory-efficient design allows SNIPER to benefit from Batch Normalization and larger batch-sizes on a single GPU. AutoFocus, the inference component, employs a coarse-to-fine approach, processing only regions likely to contain small objects at finer scales using 'FocusPixels' to generate compact 'FocusChips'. The tool supports half-precision training with no loss in accuracy and offers fast inference speeds, making it suitable for advanced computer vision research and development.

floatz provides AI-driven scientific due diligence specifically for biotech and life-science investors. The platform is designed to surface hidden scientific risks by evaluating the validity of claims and evidence within scientific literature. It connects fragmented research papers and identifies contradictions that might otherwise go unnoticed. By leveraging AI, floatz helps investors make more informed decisions, ensuring they understand the scientific landscape and potential pitfalls before committing capital to new ventures. This proactive approach helps mitigate investment risks and enhances the due diligence process for complex scientific projects.

European Language Technology serves as the central communication hub for two significant European Commission-funded projects: the European Language Grid (ELG) and European Language Equality (ELE). These initiatives are dedicated to advancing language technology across Europe, with a primary goal of achieving digital language equality by 2030. The platform provides information on the projects' objectives, consortium partners, and contact details. It also offers updates, events, and open calls, which users can stay informed about by subscribing to the ELT newsletter. The site highlights the funding received by both ELG and ELE from the European Union's Horizon 2020 research and innovation programme.

The Ellison Scholars program, offered by the Ellison Institute of Technology (EIT), provides full funding for undergraduate and graduate studies at the University of Oxford. This initiative aims to cultivate global technology innovators and leaders by immersing students in cutting-edge research and development. Scholars contribute to innovation projects across critical areas such as health, AI, climate change, and food security. The program emphasizes interdisciplinary collaboration, mentorship, and project-based learning within EIT's world-class research environment, preparing students to tackle humanity's most enduring challenges.

RejuveAI is a decentralized platform empowering longevity research and personalized insights through cutting-edge AI innovations. It aims to democratize longevity globally by allowing users to take control of their health data and harness its earning potential. The Rejuve Longevity App enables users to monitor essential health metrics, enhance their lifespan, and accumulate RJV tokens. These tokens unlock exclusive discounts on travel, supplements, medical tests, and longevity therapies. Rejuve unites researchers, clinics, and data enthusiasts in a decentralized collaboration to combat aging, ensuring innovative outcomes are affordable and accessible. Its technology uses revolutionary AI for human body analysis, crafting intricate, dynamic models of bodily functions, enriched with data from participants and various biological and medical databases. The platform also supports decentralized N-of-1 trials to generate personalized evidence on longevity optimization strategies.

Volatile AI specializes in molecule and volatile organic compound analytics, offering field instrumentation and software for end-users. The company focuses on volatilomics fingerprinting, enabling rapid chemical analysis without complex sample preparation. Their technology includes accessible gas chromatography and custom portable instruments designed for various applications such as biopharmaceutical fermentation monitoring, asphalt variant profiling, and whiskey maturation monitoring. Volatile AI positions itself as a leading research company in this field, providing solutions for understanding molecular composition outside of traditional lab settings.

RSN, or Residual Steps Network, is an open-source project developed for multi-person pose estimation, recognized as an ECCV 2020 Spotlight paper. This PyTorch-based realization won the COCO 2019 Keypoint Challenge and received the Best Paper Award. The core methodology involves aggregating intra-level features to achieve delicate local representations, crucial for precise keypoint localization. Additionally, it incorporates an efficient attention mechanism called Pose Refine Machine (PRM) to further enhance keypoint accuracy. RSN has demonstrated state-of-the-art results on both COCO and MPII benchmarks without relying on extra training data or pre-trained models. It offers various model configurations, including RSN-18, RSN-50, and RSN-101, with detailed performance metrics available for different datasets and input sizes.

Carta Healthcare offers an AI platform designed for hospitals and health systems to manage enterprise clinical data efficiently. Utilizing a unique 'Hybrid Intelligence' approach, which combines AI with expert human abstractors, the platform aims to drastically reduce costs, supercharge efficiency, save time, and enhance data quality. Key components include Voyager, an AI data platform; Lighthouse, an AI-assisted abstraction tool; and Navigator, for clinical data intelligence. It also provides expert abstraction services for flexible outsourcing or staff augmentation, helping organizations improve IRR scores and accelerate research through rapid data analysis and cohort selection for clinical trials.

North Tech is an AI research and development company focused on creating proprietary AI models and next-generation solutions. They bridge the gap between cutting-edge research and practical applications, developing industry-specific AI systems tailored to unique challenges. North Tech addresses the specific needs of sectors such as telecom, finance, tourism, insurance, and B2B services. Their mission is to continuously push the boundaries of AI innovation, providing advanced intelligence solutions that transform conversations into actionable insights for businesses.

Nētly AI is a technology company dedicated to advancing the quality of living through modern technological research. The company specializes in developing software solutions designed to empower individuals and organizations to effectively leverage various networks. Nētly AI emphasizes exploratory research, aiming to discover innovative applications of AI and technology that can lead to tangible improvements in daily life and societal well-being. The company fosters a culture of creativity and collaboration within its team, believing these values are crucial for pushing the boundaries of technological innovation.

ROKIN is an AI-driven platform specifically designed as a search engine for engineers. It assists in discovering tailored solutions for complex technical challenges by facilitating technology scouting and monitoring. The tool provides insights into hard-to-find technologies, aiming to accelerate research and innovation processes. By offering a specialized search experience, ROKIN helps engineers efficiently identify relevant market information and stay updated on technological advancements, ultimately streamlining their problem-solving and development workflows.

The Fin AI is a global academy dedicated to advancing trustworthy financial AI through community, education, research, and incubation. It aims to bring and educate the next generation of financial AI researchers and talents through courses, study groups, and hands-on learning. The initiative also drives rigorous and transparent research with open benchmarks, models, and agent systems. Furthermore, The Fin AI supports early-stage teams by transforming ideas into impactful financial AI solutions with guidance, resources, and real-world use cases. It fosters a strong community of researchers, engineers, and financial experts to bridge AI and finance.

6G-TWIN is a pioneering research initiative dedicated to establishing the foundational AI-native reference architecture for future 6G systems. The project's core innovation lies in integrating Network Digital Twins (NDT) as a central mechanism for end-to-end, real-time optimization, management, and control of highly dynamic and complex network environments. This involves creating real-time digital replicas of physical network infrastructure, providing a sandbox for training AI models and testing various scenarios before deployment. Key outputs include a flagship publication detailing a reference functional architecture for NDTs, designed to guide AI-driven, programmable, and distributed 6G networks. The project also focuses on advancing 6G security through standardization contributions and will test its solutions via teleoperated driving and energy savings demonstrators.

AI For Collective Intelligence (AI4CI) is a UKRI National AI Research Hub, bringing together universities from all four constituent nations of the UK and over forty stakeholder partners from academia, government, charities, and industry. This applied, interdisciplinary effort aims to develop novel AI solutions for collective intelligence on a national scale. The Hub specifically targets key societal challenges in areas such as Environmental Intelligence, Financial Stability, Healthcare Ecosystems, Human-Centered Design, Infrastructure and Governance, Pandemic Resilience, and Smart City Design. AI4CI is dedicated to building research capacity, capability, and community in AI for collective intelligence across the UK, offering opportunities for funding and team collaboration.

Exposys Data Labs is an AI research and development company dedicated to solving significant global challenges through technology. Their projects span various domains, including energy, where they aim to create sustainable solutions like BRICK energy from solar and wind. In defense, they've developed surveillance drones with shooting capabilities. For domestic use, they envision humanoid robots to assist with household tasks. They also focus on blockchain for secure record-keeping, medical chips for health monitoring, and E-Security with behavior-pattern-based authentication. The company emphasizes AI and ML projects, including advanced analytics, predictive analysis, spam filters, product recommenders, and fraud detectors.

ELSA – European Lighthouse on Secure and Safe AI is a growing network of excellence dedicated to spearheading research in foundational safe and secure AI methodologies. Building on the ELLIS network, ELSA unites European experts in technical robustness, privacy-preserving techniques, and human agency and oversight. The initiative also collaborates with research and industry experts across six key application sectors for safe and secure AI. ELSA provides a Benchmarks Platform for use cases, data, metrics, and competitions, fostering collaboration and scientific advancement in the field. It also offers educational resources, mobility programs, and hosts various events and workshops.