Robin Cole

icon indicating an email [email protected]

icon company @earthdaily icon location London, UK icon location Tackling the worlds toughest challenges with AI & ML applied to satellite imagery

Results 436 comments of Robin Cole

Can't get version 2 to run

hmm not sure

Checkout model maker

Pretty straight forward API for model training. Perhaps add to docs

Add transfer learning example for classification

Added in https://github.com/robmarkcole/tensorflow-lite-rest-server/commit/047cce239d60a650db58821a62f3391a5a71c247 but need refinement, models are LARGE

Try out efficient net

Turns out you need to create the model before using it, but not clear so sent an email to Manoj Your model needs to be qunatized. Try the quantized model...

Warnings while running the test scripts

I don't mean to hijack this issue, but I noted another issue where the user is not using a GPU even. I can confirm Google Colab can be used for...

Support face recognition

maybe, could get complicated

Support face recognition

https://docs.aws.amazon.com/rekognition/latest/dg/faces.html

response not used by save_image

Also camera_entity not used

Official integration

seems there might be a major refactor of image processing coming which is blocking new integrations in the meantime

Support text service

This might require quite a different config since we potentially want to extract text from multiple ROI in a single image and/or do pattern matching with regex. Consider standalone addon