Skip to content

Latest commit

 

History

History
111 lines (60 loc) · 3.92 KB

Links.md

File metadata and controls

111 lines (60 loc) · 3.92 KB

Spoiler alert - IBM approach to our problem

IBM Solution


Miro Board for Planning

Milestones and Challenges Board

Stakeholder presentation

Capstone project - Building Damage


Databases for pre-trained models (PDF 22-Image_Modeling)

1- Tensor Flow Hub

2- Keras Applications

3- PyTorch Torchvision Models

4- GitHub- Pretrained Models for PyTorch

5- GitHub-ONNX

6- GitHub-DeepSet.ai

7- Awesome Opensource List

Trained model that was used for camping project

SegFormer mit-b0

Repo Links for Pretrained models

MM Segmentation

SegFormer Original Repo

SegFormer mitb5

All SegFormer Hubs


Google Colab

Colab Websites

Marcel's brief info:
The runtime is in the free version max. 12 hours, might end if you're "inactive" (results like variables, calculations expire). Notebook files are saved to your Google Drive. But they can be saved as a regular ipynb or py file.
Moreover, GitHub can easiliy be connected to Colab, so that the file is saved to an existing repository!

Colab YouTube videos

3:10 min - Get started with Google Colaboratory (Coding TensorFlow)

5:32 min - Google Colab features you may have missed


Google cloud


Baseline model Xview2 repo

repo_baseline


Basics of Semantic Segmentation and Models

Basics of Semantic Segmentation


YouTube Videos

How to manually segment an image and use for ML

Transfer learning for Images on a small dataset


Options to Open json files easily

Json Viewer


Cloud Link for Our Project

Capstone Cloud