Middle-SENIOR COMPUTER VISION ENGINEER
The CHI Software team is not standing still. We love our job and give it one hundred percent of us! Every new project is a challenge that we face successfully. The only thing that can stop us is... Wait, it's nothing! The number of projects is growing, and with them, our team too. And now we need a Middle-Senior Computer Vision Engineer.
- Strong knowledge in computer vision fundamentals i.e. OpticalFlow, HOG, feature detection algorithms, Hough Transform, Haar Cascades, Homography, Morphology, Denoising/Deblurring, and image processing algorithms.
- Strong Python knowledge;
- Strong practical experience with DL/CV frameworks like OpenCV, PyTorch, MXNet, Tensorflow, or Keras.
- Experience with some of the well-known neural networks architectures such as Yolo, MobileNet, U-Net, R-CNN-based architectures;
- Understanding state-of-the-art CV approaches for problems like object detection/tracking, video analysis, semantic segmentation, pose estimation, optical character recognition
- Upper-intermediate level of English mandatory
Would be a plus:
- Experience with R, C++
- Experience with the following modern neural network architectures: LSTM and other RNN-based, Transformers(BERT, etc.)
- Familiarity with time-series predictive/anomaly detection analyses, natural language processing, signal processing
- Understanding SOTA approaches for machine learning problems like unsupervised / semi-supervised learning.
- Experience with the following DL frameworks: DLib, Darknet, Theano.
- Awareness of the CRISP-DM process model
- Experience with continuous integration and release management tools, preferably within the AWS platform.
- Hands-on Experience with the common architecture of MLOps system by the means of Hadoop, Docker, Kubernetes, cloud services and experience with managing production ML lifecycle
It would be time-series sensor data from sensors in the ground combined with satellite imagery to predict that a dam is about to be breached. Dataset analysis and creation, Data visualization, Data modeling.