Computer Vision
Computer Vision

Being one of the most exciting AI fields, computer vision is a multidisciplinary area that deals with intelligent processing of visual information. It is behind the scenes of fingerprint recognition and biometrics in your smartphone, automated translation from an image, automotive safety, streamlined visual inspection of mass production, and many other fascinating tasks. Here, at It-Jim, we are passionate about computer vision R&D and engineering.

Image processing
Image processing

Since any arbitrary physical parameter can be encoded and visualized as an image, proper digital image processing solutions can be really handful in projects from many domains. Indeed, visual information like images and videos is the most widely used in all businesses, which is no surprise as the human brain instantly understands and interprets it. Basically, image processing can be considered as a type of two-dimensional signal processing applied to image pixels.

Machine Learning
Machine Learning

Today, when we have long entered the era of artificial intelligence, computer vision (CV) could not stay aside from various techniques of machine learning (ML). Moreover, with a rapid progress of deep learning algorithms and, in particular, convolutional neural networks (CNNs), automatic analysis of visual information is reaching a new level.

Signal Processing
Signal Processing

From smartphones to wearable devices, from healthcare to finance, signal processing is much closer than you think. Basically, any information can be represented as a signal: speech, audio, image, video, text, stock or electricity prices, medical parameters or any other arbitrary data. The signal processing engineering, then, can be used for the extraction, interpretation and transformation of many different types of information.

1D direction estimation with a YOLO network

The modified You only look once (YOLO) network architecture that allows one-dimensional direction estimation along with classic object detection in real time, is considered in the task of street traffic surveillance from unmanned aerial vehicles. The key feature is a modified output fully connected layer with additional orientational parameters. It has been shown that this network can estimate the direction of vehicles on a custom testing dataset with photos.

Detection of Airplanes on the Ground Using YOLO Neural Network

The presented paper benchmarks the performance of state-of-the-art methods of objects detection in the particular case of airplanes on the ground identification detection in aerial images taken from unmanned aerial vehicles or satellites. There were tested two popular single-stage neural networks YOLO v.3 and Tiny YOLO v.3 based on the “You Only Look Once” approach. The considered artificial neural network architectures for objects detection has been trained and applied over the specifically created image database.