Neha Sharma

Create Python Jupyter notebooks, train geospatial AI models, develop GeoAI-enabled solutions, and draft comprehensive documentation. Lead the development and execution of test plans and test cases, ensuring the quality and reliability of code changes and bug fixes. Craft technical content for help topics, blog posts, guides, presentations, tutorials, and sample notebooks. Monitor and maintain ESRI GeoAI pretrained models hosted on ArcGIS Living Atlas of the World. Update and maintain ESRI deep learning Jupyter Notebooks on the ArcGIS Developer Website and ArcGIS AWS Enterprise portal. Spearheaded the development of a Ship Detection (RGB) deep learning pretrained model utilizing Mask-RCNN within the ArcGIS API for Python framework. Authored technical documentation on topics such as Fairness in Train Using AutoML tool, Test Time Augmentation Scales (TTA Scales), and various pretrained models. Conducted rigorous quality assurance and certification processes for ArcGIS API for Python, ArcGIS GeoAI pretrained models, ArcGIS GeoAI Timeseries AI and Spatial Statistics tools, ArcGIS GeoAI Train using AutoML tool, and Multiscale TTA Feature in ESRI TextSAM Model. Written and published blogs on ArcGIS AI Models Year in Review and Boosting GeoAI Model Performance with Test Time Augmentation.

Developed full motion video analytics web application using advanced photogrammetric techniques in drone videos. Developed geospatial deep learning models for object detection in drone videos using Pytorch MaskRCNN and instance segmentation. Developed geospatial web applications for drone video analytics and feature detection using Python Flask, HTML, CSS, Leaflet, and Folium Maps. Designed GUIs for models and algorithms using PyQt5 and Flask. Coordinated knowledge sharing sessions and organized International Women's Day event. Won Mindtree NxT Hackathon Game Changer Award 2021 for GeoAI and city cleanliness.

Handled large biodiversity geospatial datasets on Azure Databricks for uploading, processing, and analytics using GeoAI and Python programming. Implemented advanced photogrammetric techniques in drone videos to determine object geo-locations using drone metadata and SRT files. Developed geospatial deep learning models for object detection in drone videos using Pytorch MaskRCNN and instance segmentation. Developed geospatial web applications for drone video analytics and feature detection using Python Flask, HTML, CSS, Leaflet, and Folium Maps. Streamlined drone data processing by converting SRT files to CSV and JSON formats. Designed GUIs for models and algorithms using PyQt5 and Flask. Developed open-source geospatial deep learning models for object detection, custom instance and semantic segmentation using Mask RCNN, YOLOv5, UNET, TensorFlow, and Pytorch. Developed python algorithms for metadata extraction from geotagged 360 photos and projection conversion of shapefiles using Geopandas and Pyproj. Automated ArcGIS map layout and attribute table PDF generation using ReportLab and Arcpy. Conducted geospatial training sessions for non-geospatial team members. Volunteered as an online educator for underprivileged children with Angel Xpress Foundation (AXF) NGO. Contributed to GeoAI and GIS content generation for L&T-EduTech.

Developed geospatial web applications for drone video analytics and feature detection using Python Flask, HTML, CSS, Leaflet, and Folium Maps. Streamlined drone data processing by converting SRT files to CSV and JSON formats. Designed GUIs for models and algorithms using PyQt5 and Flask. Developed python algorithms for metadata extraction from geotagged 360 photos and projection conversion of shapefiles using Geopandas and Pyproj.

Assisted in teaching Geographic Information Science Foundation Studio for M.Tech (Geomatics). Conducted lab sessions, project guidance, and spatial data tutorials.

Contributed to Urban Air Quality Analysis for Ahmedabad using Sentinel-5P and MODIS data. Integrated RS, GIS, and AQ models; project awarded Best Capstone Project 2020 (Geomatics).