Use deep learning models in ArcGIS Notebooks

ArcGIS 11.5 | |  Help archive

Deep learning is a rapidly evolving method of machine learning data analysis that uses neural networks to mimic the processes of the human brain. When integrated with GIS, it can enhance data handling and provide more accurate data analysis.

There are typically three major challenges associated with machine learning algorithms: analyzing large amounts of data, insufficient computational resources, and the efficiency and accuracy of algorithms. By integrating ArcGIS Notebooks and deep learning algorithms, you can better manage these challenges.

Deep learning in ArcGIS Notebooks

The deep learning models for imagery data available in ArcGIS Notebooks fall into four major categories:

  • Object classification—Used to determine the class of a feature. For example, it can be used to determine whether a building is damaged after a natural disaster.
  • Object detection—The localization process of finding an object in an image within a bounding box. For instance, object detection can be used to detect swimming pools in an image.
  • Pixel classification—Used to assign classes to pixels within an image. For example, pixel classification can be used for land cover classification.
  • Instance classification—Integrates both object detection and pixel classification. Instance classification can be used to detect an object and to further classify it. For instance, pixel classification can be used in damage detection to not only identify damage, but to further classify the severity of the damage.
Note:

ArcGIS Notebooks also support models focusing on tabular, point cloud, and other structured datasets.

Deep learning workflows

ArcGIS Notebook Server can be used to perform the following deep learning workflows:

  • End-to-end deep learning
  • Inferencing using pretrained models
  • Fine-tuning pretrained models

End-to-end deep learning

End-to-end deep learning is a transformative method in machine learning where a single neural network is trained to perform complex tasks directly form raw input data, bypassing the need for manual feature extraction. This process involves the following steps:

  1. Data collection—Gather a large and diverse dataset relevant to the task.
  2. Data preprocessing—Clean and prepare the data for training.
  3. Model design—Choose a neural network architecture.
  4. Training—The model is trained on the dataset and the parameters are adjusted to minimize errors.
  5. Evaluation—Test the model on a separate dataset to assess the model's performance.
  6. Deployment—Implement the trained model in a real-world application.

See Example: End-to-end deep learning workflow for an example workflow.

Pretrained models

Pretrained deep learning models can streamline your geospatial workflows by eliminating the need for extensive training data and computational resources. ArcGIS offers a number of pretrained deep learning models for different tasks. These models can be downloaded from ArcGIS Living Atlas and used in ArcGIS Notebooks by uploading them as portal content.

See Example: Inferencing using a pretrained model for an example workflow.