{"id":31320603,"url":"https://github.com/dwade-eng/hermes---ai","last_synced_at":"2026-05-16T11:04:00.424Z","repository":{"id":315976172,"uuid":"1061446998","full_name":"dWADE-ENG/Hermes---AI","owner":"dWADE-ENG","description":"I'm currently developing Hermes AI, a modular deep learning framework for post-disaster flood, damage, and debris segmentation — leveraging datasets like FloodNet, RescueNet, and SpaceNet-8.  My focus is on blending satellite imagery, drone footage, and multi-head segmentation models to support FEMA critical infrastructure recovery operations. ","archived":false,"fork":false,"pushed_at":"2025-09-21T23:39:59.000Z","size":3056,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":0,"default_branch":"main","last_synced_at":"2025-09-22T01:13:36.340Z","etag":null,"topics":["colab-notebook","computer-vision","deeplabv3","drone-plugin","excel","gis","machine-learning","python3","resnet-50","uav-images"],"latest_commit_sha":null,"homepage":"https://gems-ai.io/","language":"Jupyter Notebook","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/dWADE-ENG.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2025-09-21T23:14:42.000Z","updated_at":"2025-09-21T23:40:02.000Z","dependencies_parsed_at":"2025-09-22T01:13:56.043Z","dependency_job_id":"518fb6d4-006d-43c5-951e-1f3cf40b6216","html_url":"https://github.com/dWADE-ENG/Hermes---AI","commit_stats":null,"previous_names":["dwade-eng/hermes---ai"],"tags_count":null,"template":false,"template_full_name":null,"purl":"pkg:github/dWADE-ENG/Hermes---AI","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/dWADE-ENG%2FHermes---AI","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/dWADE-ENG%2FHermes---AI/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/dWADE-ENG%2FHermes---AI/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/dWADE-ENG%2FHermes---AI/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/dWADE-ENG","download_url":"https://codeload.github.com/dWADE-ENG/Hermes---AI/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/dWADE-ENG%2FHermes---AI/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":279017381,"owners_count":26086052,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","status":"online","status_checked_at":"2025-10-13T02:00:06.723Z","response_time":61,"last_error":null,"robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["colab-notebook","computer-vision","deeplabv3","drone-plugin","excel","gis","machine-learning","python3","resnet-50","uav-images"],"created_at":"2025-09-25T16:51:19.741Z","updated_at":"2025-10-14T00:13:15.189Z","avatar_url":"https://github.com/dWADE-ENG.png","language":"Jupyter Notebook","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Hermes---AI\nI'm currently developing Hermes AI, a modular deep learning framework for post-disaster flood, damage, and debris segmentation — leveraging datasets like FloodNet, RescueNet, and SpaceNet-8.  My focus is on blending satellite imagery, drone footage, and multi-head segmentation models to support FEMA critical infrastructure recovery operations. \n\n# 🛰️ Hermes AI\n*Geospatial Damage Intelligence for a New Era of Disaster Response*\n\n**Hermes AI** is a modular deep learning framework built for **UAV-based damage mapping**, **flood segmentation**, and **infrastructure resilience assessment** in post-disaster scenarios. Leveraging high-resolution aerial imagery and satellite data, Hermes integrates multi-head semantic segmentation models to automate damage detection at scale for use by FEMA, public agencies, and disaster management contractors.\n\nHermes is the flagship project of **GEMS AI**, a Florida-based emergency tech company focused on applying computer vision, AI, and GIS to real-world disaster relief operations.\n\n---\n\n## 🌪️ Why Hermes?\n\nIn the critical hours after a hurricane, flood, or earthquake, the ability to rapidly map structural damage and flooded roads can mean the difference between chaos and coordinated response.\n\nTraditional damage assessments are:\n- **Slow**: Manual field inspections can take weeks.\n- **Subjective**: Paper-based PDA systems vary across jurisdictions.\n- **Inaccessible**: Delays in data centralization and sharing.\n\nHermes solves this by:\n- Using UAVs and satellite imagery to **detect damage autonomously**.\n- Integrating **multi-class segmentation models** to label water, debris, roofs, and structural failure.\n- Fusing with GIS to allow **real-time, map-based visualization** for response teams and agencies.\n\n---\n\n## 🧠 Core Architecture\n\nHermes AI uses a multi-task, multi-head architecture where each model performs a specific geospatial task:\n\n| Sub-Model        | Task                                      | Dataset           | Backbone         |\n|------------------|-------------------------------------------|-------------------|------------------|\n| `FloodNetHead`   | Flood segmentation                        | FloodNet, SpaceNet8 | DeepLabV3+ (ResNet-101) |\n| `DamageNetHead`  | Building damage classification (none/minor/major/destroyed) | xBD, RescueNet    | DeepLabV3+       |\n| `DebrisNetHead`  | Road debris segmentation                  | Custom UAV labels | UNet             |\n| `FireNetHead`    | Burn scar detection (WIP)                 | FireNet, MODIS    | EfficientNet     |\n\nAll models are modular and can be run **individually or as an ensemble**, depending on the deployment scenario.\n\n---\n\n## 🧪 Key Features\n\n✅ Multi-task segmentation (flood, damage, debris)\n\n✅ Pre-/post-event change detection (using dual inputs)\n\n✅ UAV + Satellite compatibility (.tif, .jpeg, .png)\n\n✅ Google Colab A100-ready training notebooks\n\n✅ Export to GeoTIFF and ArcGIS integration\n\n✅ Designed for on-site field operation via local inference or cloud-streamed inputs\n\n---\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdwade-eng%2Fhermes---ai","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fdwade-eng%2Fhermes---ai","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fdwade-eng%2Fhermes---ai/lists"}