give papa some data // computer vision applied to the world (for science and fun)
definitely not skynet. definitely.
seattle/oregon/cali wildfire shit
- create crawlers for one stream of timestamped images of "open views"
- save and compress images with metadata
- correlate gps and surveillance
- run labeled dataset through basic cnn model (and then the full
vizinet
model) - 🕳️
high-level research/motivation below. dunno if we'll get there
- social distancing compliance via covid-19
formulating earth air quality purely as a computer vision problem. sources of pm2.5 (particulate matter 2.5):
- wildfires
- factories
- power plants
- cars
- cow burps and farts (yum)
this has benefits primarily for developing countries in place of maintaining expensive equipment for the same task. only proof point is making it more accurate so the tradeoff becomes clear and we can start using citizens as sensors for metrics that improve decision making for the good of the same populus
hypothesis: understanding exponential attenuation caused by particulate matter can help depth estimation systems and automonomous vehicles driving in adverse weather (not just on earth). potential for transfer learning here, but i could be wrong
ideal collection locations: seattle (today, for wildfires), beijing, india
- vizinet app (human, not so automated)
- https://www.insecam.org/ - "world biggest directory of online surveillance security cameras"
- https://www.earthcam.com/usa/washington/seattle/?cam=seattledowntown#gall-29
- https://www.geocam.ru/en/
- https://worldcam.eu/search?q=seattle&as_values_result=
- https://www.skylinewebcams.com/en/webcam/china/hong-kong/hong-kong/china-hong-kong.html
- https://www.tour-beijing.com/
- https://www.webcamtaxi.com/en/usa/california/homewood-mountain-resort.html
- http://www.wasar.org/webcams/
- https://www.esbnyc.com/about/live-cam
- https://www.purpleair.com/map#1/25/-30
- that one dataset i got access to from those beijing mobile app paper authors
- already crawled and correlated tour-beijing images
- https://earthdata.nasa.gov/earth-observation-data/near-real-time/hazards-and-disasters/air-quality - satellite images and air quality; neat idea
- https://github.com/awesomedata/awesome-public-datasets#climate-weather - some data to cross-reference here, probably
- https://www.youtube.com/watch?v=M2Y9n6fhoxI - build a cheap particular matter sensor for under $100. could be neat to combine with existing raspberry pi 1080p camera.
- https://fischcheng.github.io/post/smoke_seattle/
- https://github.com/topics/image-crawler
- https://github.com/CDCgov - for the covid stuff
- https://github.com/topics/surveillance
- https://github.com/topics/wildfire - wildfire detection and prevention stuff
- https://github.com/k4cg/nichtparasoup/blob/3.0-dev/python-package/README.md - this sort of random dashboard idea is cool
- https://github.com/matiasraisanen/insecrawl - insecam crawler
- https://github.com/justrandomwebcams/totalynothackedijokeyounot - an insecam dump of links (w/o crawler code)
- https://github.com/omic-ai/jailbreak - shameless plug
- https://github.com/scrapy/scrapy - pretty popular
- https://github.com/TransparencyToolkit/Harvester
- https://github.com/mathdroid/covid-19-api
- https://github.com/TransparencyToolkit/Surveillance-Research-Data
- https://code.google.com/archive/p/surveillancesaver/ - looks like a good source/idea
- https://www.300dollardatarecovery.com/surveillancesaver-live-video-screen-saver-broadcasting-from-around-the-world/
- https://github.com/Pamoi/webcam-crawler
- https://github.com/cloudviz/agentless-system-crawler
- https://github.com/COVID19Tracking
- https://github.com/SuperBuker/CamHell
- https://github.com/yash-1708/Surveillance-crawler
- https://gist.github.com/jeffcrouse/7797275
- https://github.com/hakluke/hakrawler
- https://github.com/mykle1/MMM-EARTH-Live - from the iss, just for fun
- https://github.com/damien911224/theWorldInSafety - ok idea, not sure how much they actually did this. wish there was a paper
- https://github.com/Anukriti2512/Air-Quality-Prediction
- https://github.com/grtvishnu/Air-Pollution-Prediction-and-Forecasting
- https://github.com/ginberg/mlcapstone
- https://github.com/spatial-computing/air-quality-prediction-scala
- https://github.com/arnavbansal1/SatellitePollutionCNN
- https://github.com/arnavbansal1/SatellitePollutionCNN
- https://github.com/curiousest/predict-AQI
- https://github.com/curiousest/predict-AQI
- https://github.com/vikmreddy/deep-air
- https://github.com/luna983/air-quality-machine-learning
- https://github.com/ashishpatel26/Real-time-ML-Project
- https://github.com/txytju/air-quality-prediction
- https://github.com/Alro10/deep-learning-time-series