search

kdmayer / PV_Pipeline

PyTorch models and pipeline developed for "DeepSolar for Germany". For reference, the paper can found at https://ieeexplore.ieee.org/document/9203258
MIT License
24 stars 6 forks source link

About trained checkpoint #2

Closed yjhong89 closed 3 years ago

yjhong89 commented 3 years ago

Dear authors of deep solar for germany,

Is there a trained checkpoint for classifying solar panels ?

I saw 'checkpoint_path' in config.yml file, but there isn't any checkpoint file in this directory.

Thank you in advance.

Regards, YJHong.

kdmayer commented 3 years ago

Hi YJHong,

Thank you for reaching out!

I will bring up your request in our upcoming research group meeting (November 16th) and get back to you once I had a chance to speak with my professor.

Best regards,

Kevin

kdmayer commented 3 years ago

Hi YJHong,

I have uploaded both model checkpoints to the repository.

DeepSolar_GoogleMaps_classification.tar should have a precision of 92% and a recall of 98% on aerial imagery provided by Google Maps Static API (zoom level=21, GSD of 5 cm/pixel)

DeepSolar_openNRW_classification.tar should have a precision of 64% and a recall of 87% on aerial imagery provided by the openNRW portal (GSD of 10 cm/pixel)

Best,

Kevin

yjhong89 commented 3 years ago

@kdmayer

Thank you a lot!

YJ Hong.