Let us develop your customized Computer Vision solution
We develop tailor-made models, designed and trained to meet your business needs, based on the latest advances in scientific research.
Let's discuss your projectLet's discuss your project
Computer Vision

What is Computer Vision?

Computer Vision is a field of artificial intelligence that involves the analysis of images and videos. It enables automated systems to perform complex tasks, such as object recognition, motion detection or text reading (OCR).
Application examples
Computer vision covers a wide range of techniques for analyzing and understanding images in different forms. The examples below are just a few of the countless possible applications: from simple object recognition to the detailed interpretation of complex scenes.
undefined
Object detection consists in automatically recognizing and locating an element in an image. This operation often provides greater precision than classification alone: it generates a bounding box around the detected object. This visual localization can be crucial for certain applications, such as video surveillance or in-store product identification.
Semantic segmentation aims to automatically assign a category - or mask - to each pixel in an image, enabling different areas to be delineated very precisely. A segmentation model is particularly relevant in fields such as precision mapping for navigation systems or fine tissue analysis in the medical sector.
undefined
Pose estimation involves analyzing and locating key points on the human body, such as joints, in order to establish posture and deduce movements. This approach is proving crucial in a number of fields, including sports training, functional rehabilitation and real-time animation in the entertainment sector.
undefined
Visual Question Answering involves automatically answering a question posed from an image, also requiring a combination of computer vision and natural language processing. In certain contexts, thanks to targeted queries, a Visual Question Answering model is capable of extracting more precise information than that provided by Image Captioning.

Our expertise for your project

Service Picture
We leverage the most powerful pre-trained Deep Learning models on the market, such as ResNet classifiers, the YOLO detection model or the SAM segmentation model. Depending on your product or business need, we select the most appropriate model and fine-tune your data to achieve maximum performance.

Does your vision model need to be integrated into a business application? Accessible to hundreds or thousands of users, while optimizing hosting costs? Does it need to be resized for an embedded system with limited computing capacity? Galadrim can support you at every stage of your project, with engineers experienced in the integration and deployment of complex AI models.

Team

We are a team of 30 engineers and consultants passionate about artificial intelligence and the practical application of the latest advances in research. Our mission: to democratize and simplify the use of AI for all our customers.
Let's discuss your project
Founders
Pablo, Partner & Head of AI
A graduate of HEC, Pablo is in charge of Galadrim's AI business development and advises our customers on the strategic aspects of their projects. Pablo previously worked as a mathematics teacher in higher education.
View LinkedIn profile
Benjamin, Partner & CTO AI
An engineer from the École Polytechnique and the Corps des Mines, Benjamin supervises the technical team and advises our customers on the technologies best suited to their needs. He entered the world of entrepreneurship after a career in senior civil service, and has held a number of technical leadership roles.
View LinkedIn profile

Our achievements

Image
Guillin Group
Packaging pattern recognition
Development and fine-tuning of a model to recognize one of Guillin's 25,000 packaging references from an image.
  • Development of a mobile application dedicated to data collection and operational use of the model by the Group's 250 sales staff.
  • Adaptation and fine-tuning of pre-trained classification and detection models.
  • Deployment of the application and models on a suitable infrastructure.
Video
Gozoki
Theft detection
Design and development of a real-time accident and theft detection system based on video surveillance of autonomous stores.
  • Design of the complete system and supervision of service providers responsible for hardware infrastructure (camera installation, etc.).
  • Adaptation and fine-tuning of pre-trained pose detection and estimation models.
  • Development of a web application enabling security personnel to view alerts in real time and intervene remotely in stores.
  • Orchestration of video flows Deployment of the solution on a suitable infrastructure.

Our technologies

TensorFlow
TensorFlow
PyTorch
PyTorch
OpenCV
OpenCV
Tesseract OCR
Tesseract OCR
Segment Anything
Segment Anything
GPT-4o
GPT-4o
Do you have a project?
Start the conversationStart the conversation