Open salmaelgayar opened 7 months ago
The model is not fine-tuned to handle X-ray Images. It would require further training (on X-ray images) to be able to do so.
Thank you for your reply. Actually, I solved the problem, did what you said, and trained the medical x-rays on the model. but I couldn't reach high accuracy with the validation dataset (80%). Any recommendations for the best fine-tuned neural network layer to be added to this architecture to get the best caption validation results?!
Hello Thank you for sharing the amazing code! I tried to run this code on different X-ray data sets for medical captioning. Unfortunately, I always get the same word for caption: "the the the..." for any input image Do you have any suggestions as to why this happened?