API Reference

Community

Bluesight

Contact Us

Run inference of trained model on your data.

Run Trained Model Inference

Order of bands in pixels array. Used to retrieve means, stds and wavelengths. Pick from 'blue', 'green', 'red', 'rededge1', 'rededge2', 'rededge3', 'nir', 'nir08', 'swir16', 'swir22'.

Bands

gsd's of each band in images list (should be the same across all bands)

3D float array (number of bands, h, w) representing an image.

Pixels

One of ['sentinel-2-l2a', 'landsat-c2l1', 'landsat-c2l2-sr', 'naip', 'linz', 'sentinel-1-rtc']. Used to retrieve means, stds and wavelengths across bands. If not provided, means and stds will be calculated from list of images, and default wavelengths will be used

Platform

(lat, lon) coordinate of the center of an image. Doesn't have a huge impact on model output.

Point

Timestamp when image was taken. Doesn't have a huge impact on model output.

Timestamp

Bands' wavelengths. Should be provided only if `platform` is `None`, because when `platform` is provided wavelengths will be loaded from config. If `platform` is `None` and wavelengths are not provided, default values from `sentinel-2-l2a` will be used: {'blue': 0.493, 'green': 0.56, 'red': 0.665, 'rededge1': 0.704, 'rededge2': 0.74, 'rededge3': 0.783, 'nir': 0.842, 'nir08': 0.865, 'swir16': 1.61, 'swir22': 2.19}

Wavelengths

Image

Location

Message

Error Type

ValidationError

InferenceData

Response Run Trained Model Inference Inference Post

HTTPValidationError

Bluesight is an AI platform for geospatial data. We offer hosted inference of [Clay foundation model](https://github.com/Clay-foundation/model) and a fine-tuning API for downstream tasks. All endpoints and support on the platform are free and will stay so through mid August. Please do get in touch at hello@bluesight.ai if you have questions or suggestions.

Introduction

Start training satellite images classification models using Bluesight API

Train a ship classifier

Train a forest fire classifier

Start training satellite images segmentation models using Bluesight API

Train a land segmentation

Creates a training job which begins the process of creating a new model from a given dataset.

Response includes details of the enqueued job including job status and the name of the fine-tuned models once complete.

Create Training Job

Retrieve Training Job

Cancel Training Job

Get Embeddings With Images

Get Embeddings With Coordinates

Get Embeddings With Text

Upload a file to the Bluesight storage bucket.

The file may be used for [creating a training job](https://docs.bluesight.ai/api-reference/training-jobs/create-training-job) or [inference](https://docs.bluesight.ai/api-reference/inference/run-trained-model-inference) (TBD).

File is expected to be in HDF5 format. Check guides page for info on how to create a valid file and upload it (TBD).

Training Jobs

Inference

Embeddings

Files

Models

Run Trained Model Inference

Body

Response