I have a small project to identify different plants and flowers in my garden so I'm looking for best computer vision model or even a combination of LLMs out there. I found several computer vision models perform well in plant identification. Here are some top contenders to consider, depending on your specific needs:
1. Convolutional Neural Networks (CNNs):
- These are the current state-of-the-art for plant identification.
- Popular pre-trained CNN options include:
- ResNet: Known for its accuracy and efficiency, especially variants like ResNet-50 or the customized ResNet26 for plant images.
- MobileNetV2: A lighter-weight option ideal for mobile applications due to its faster processing speed with good accuracy.
2. Other Models:
- Bag-of-Words (BoW) model with Scale-Invariant Feature Transform (SIFT): This approach extracts keypoint features from the image and uses them for classification. While not the most advanced, it can be a good option for smaller datasets.
Choosing the best model depends on factors like:
- Dataset size and variety: CNNs generally perform better with extensive training data.
- Computational resources: ResNet might be more demanding than MobileNetV2.
- Application requirements: If you need a mobile app, MobileNetV2's efficiency might be crucial.
Here are some additional resources to help you decide:
- Research paper comparing models: A Novel Computer Vision Model for Medicinal Plant Identification Using Log-Gabor Filters and Deep Learning Algorithms: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9532088/
- Deep Learning for Plant Identification: Deep Learning for Plant Identification in Natural Environment: https://www.worldscientific.com/doi/full/10.1142/S1469026822500055
- Plant Species Identification Using Computer Vision Techniques: [Plant Species Identification Using Computer Vision Techniques: A Systematic Literature Review ON Springer link.springer.com]
Use the comment below to post other ideas. I will share the website or app after my research and testing when it's completed.
More Computer Vision or Machine Learning projects at MachineLearn.com
First posted at QUE.com