Martin Tomov
commited on
Commit
•
f278fa4
1
Parent(s):
3f208ba
rm old readme
Browse files
README.md
DELETED
|
@@ -1,55 +0,0 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
---
|
| 4 |
-
|
| 5 |
-
# InsectSAM: Insect Segmentation and Monitoring
|
| 6 |
-
|
| 7 |
-
<p align="left">
|
| 8 |
-
<a href="" rel="noopener">
|
| 9 |
-
<img width=200px height=200px src="https://i.imgur.com/hjWgAN9.png alt="Project logo"></a>
|
| 10 |
-
</p>
|
| 11 |
-
|
| 12 |
-
## Overview
|
| 13 |
-
|
| 14 |
-
InsectSAM is an advanced machine learning model tailored for the https://diopsis.eu camera systems and https://www.arise-biodiversity.nl/, dedicated to Insect Biodiversity Detection and Monitoring in the Netherlands. Built on Meta AI's `segment-anything` model, InsectSAM is fine-tuned to be accurate at segmenting insects from complex backgrounds, enhancing the accuracy and efficiency of biodiversity monitoring efforts.
|
| 15 |
-
|
| 16 |
-
## Purpose
|
| 17 |
-
|
| 18 |
-
This model has been meticulously trained to identify and segment insects against a variety of backgrounds that might otherwise confuse traditional algorithms. It is specifically designed to adapt to future changes in background environments, ensuring its long-term utility in the DIOPSIS / ARISE project.
|
| 19 |
-
|
| 20 |
-
## Model Architecture
|
| 21 |
-
|
| 22 |
-
InsectSAM utilizes the advanced capabilities of the `segment-anything` architecture, enhanced by our custom training on an insect-centric dataset. The model is further refined by integrating with GroundingDINO, improving its ability to distinguish fine details and subtle variations in insect appearances.
|
| 23 |
-
|
| 24 |
-
## Quick Start
|
| 25 |
-
|
| 26 |
-
### Prerequisites
|
| 27 |
-
|
| 28 |
-
- Python
|
| 29 |
-
- Hugging Face Transformers
|
| 30 |
-
- PyTorch
|
| 31 |
-
|
| 32 |
-
### Usage
|
| 33 |
-
|
| 34 |
-
#### Install
|
| 35 |
-
``` bash
|
| 36 |
-
!pip install --upgrade -q git+https://github.com/huggingface/transformers
|
| 37 |
-
!pip install torch
|
| 38 |
-
```
|
| 39 |
-
#### Load model directly via HF Transformers 🤗
|
| 40 |
-
``` bash
|
| 41 |
-
from transformers import AutoProcessor, AutoModelForMaskGeneration
|
| 42 |
-
|
| 43 |
-
processor = AutoProcessor.from_pretrained("martintmv/InsectSAM")
|
| 44 |
-
model = AutoModelForMaskGeneration.from_pretrained("martintmv/InsectSAM")
|
| 45 |
-
```
|
| 46 |
-
|
| 47 |
-
### Notebooks
|
| 48 |
-
|
| 49 |
-
Three Jupyter notebooks are provided to demonstrate the model's capabilities and its integration with GroundingDINO:
|
| 50 |
-
|
| 51 |
-
- **InsectSAM.ipynb**: Covers the training process, from data preparation to model evaluation.
|
| 52 |
-
- **InsectSAM_GroundingDINO.ipynb**: Demonstrates how InsectSAM is combined with GroundingDINO for enhanced segmentation performance.
|
| 53 |
-
- **Run_InsectSAM_Inference_Transformers.ipynb**: Run InsectSAM using Transformers.
|
| 54 |
-
|
| 55 |
-
Check out the notebooks on RB-IBDM's GitHub page - https://github.com/martintmv-git/RB-IBDM/tree/main/InsectSAM
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|