A Wide Spectrum of Applications
We cover a large panel of Computer Vision use-cases
and meet the needs of various industries
Image & Video Classification, Prediction
Training a convolutional deep neural network (CNN) to classify images is a powerful Deep Learning technology allowing to predict the category of an image from large labeled datasets. Convolutional networks extract by themselves complex features at different scales from an input image and finally output its probability to belong to a specific class.
Medical
Pharma
AI medical image analysis, diagnosis, monitoring, quantification, ultrasound, optical images, MRI
Biology
Cosmetics
AI skin analysis, biological tissues & cells images, visualization, quantification, statistics
Aerospatial
AI satellite or drone image processing, feature and change detection, image registration
Games
Augmented Reality
AI scene detection on smartphones and tablets, face tracking, augmented reality applications
Object Detection, Segmentation, Tracking
Detecting the position of a specific object in an image, drawing its contours, or even tracking the object contours in a video, are famous Computer Vision tasks. Providing a large dataset of annotated images by experts, convolutional deep neural networks address such tasks with very high performance. In some specific cases such like Medical applications, smart annotations tools using handcrafted Computer Vision make the annotations faster and more efficient.
Image Matching, Registration, Stitching
You need to align different images of the same scene ? For instance to perform motion correction or to create panoramas by stitching images ? You face an image matching-registration problem. Modern Deep Learning approaches consist of extracting robust features that describe image regions and own invariance properties against various geometric or modality changes.
Image & Video Enhancement, Colorization, Generation
Generative neural networks (GANs) have opened new horizons in Computer Vision and Image Processing. Providing examples to the machine, trained models are able to enhance quality of images such as image denoising or deblurring. Converting an image towards an other modality or even generating new fashioned images using the artificial intuition of the machine represent very promising Deep Learning image and video applications.