Hello!
I made a roboflow model for my project few months ago and i love how easy it is.
I was wondering, since my prediction % isn’t good enough’ if it’s possible to “copy” the current model and edit it by adding more 100s of pictures
when i say copy it, is in order to prevent destroying the old model by mistake
and try adding more pictures for better results
Absolutely, you can improve your model’s performance by adding more images to your dataset. Roboflow allows you to add more images to your existing dataset and retrain your model. This process is part of what is known as Active Learning.
However, to ensure you don’t overwrite your existing model, you might want to create a new version of your dataset with the additional images. This way, you can train a new model on this expanded dataset while keeping your old model intact.
Remember, the number of images you need to train a model can vary widely. It can work with 100 images, 500 images, or with 10,000. It just depends on what you are doing and the level of accuracy required for you to use your model (source).
Also, Roboflow recommends a data-centric approach to computer vision. It’s often easier to adapt, update, and change the data training your model than editing the model itself (source).
Please note that while you can add more images and retrain your model, the performance of the model also depends on the quality of the images and the annotations, the diversity of the dataset, and the model architecture itself.
Hi! I’m Roboflow’s AI powered bot. This is an automated response powered using AI with knowledge from our docs, blog, and forum. If this solves your issue, please mark it as the solution using the checkbox icon. If this doesn’t solve your issue, please follow up with what you’ve tried/why, and the community will continue to chime in as usual.
You can add more data and Generate a new dataset Version. This will be a unique Version to then train a model. All of your previous dataset Versions and Models will not be impacted. You will have multiple Models and Versions to toggle between or use.