Data Scientist - Computer Vision
- New Delhi
- Permanent
- Full-time
- Develop tools, APIs and pretrained models for geospatial AI
- Integrate ArcGIS with popular deep learning libraries such as PyTorch
- Develop APIs and model architectures for computer vision and deep learning applied to geospatial imagery
- Author and maintain geospatial data science samples using ArcGIS and machine learning/deep learning libraries
- Curate and pre/post-process data for deep learning models and transform it into geospatial information
- Perform comparative studies of various deep learning model architectures
- 2 to 6 years of experience with Python, in data science and deep learning
- Self-learner with coursework in and extensive knowledge of machine learning and deep learning
- Experience with Python machine learning and deep learning libraries such as PyTorch, Scikit-learn, NumPy, Pandas
- Expertise in one or more of the following areas:
- Traditional and deep learning-based computer vision techniques with the ability to develop deep learning models for computer vision tasks (image classification, object detection, semantic and instance segmentation, GANs, super-resolution, image inpainting, and more)
- Convolutional neural networks such as VGG, ResNet, Faster R-CNN, Mask R-CNN, and others
- Transformer models applied to computer vision
- Expertise in 3D deep learning with Point Clouds, meshes, or Voxels with the ability to develop 3D geospatial deep learning models, such as PointCNN, MeshCNN, and more
- Experience in data visualization in Jupyter Notebooks using matplotlib and other libraries
- Experience with hyperparameter-tuning and training models to a high level of accuracy
- Bachelor's in computer science, engineering, or related disciplines from IITs and other top-tier engineering colleges
- Existing work authorization for India
- Experience applying deep learning to satellite or medical imagery or geospatial datasets
- Familiarity with ArcGIS suite of products and concepts of GIS