search

KaihuaTang / Scene-Graph-Benchmark.pytorch

A new codebase for popular Scene Graph Generation methods (2020). Visualization & Scene Graph Extraction on custom images/datasets are provided. It's also a PyTorch implementation of paper “Unbiased Scene Graph Generation from Biased Training CVPR 2020”
MIT License
1.03k stars 228 forks source link

Finetune on custom dataset? #174

Open DesaleF opened 2 years ago

DesaleF commented 2 years ago

Hi, Dear Author! First thank you for sharing your work. I explore the code and also tested the model on random images. I found your model very useful for my project. However, I have object labels and relationship(predicates) which doesn't exist in visual-gnome dataset. So I decided to finetune your model using my custom dataset. Can you help me with the steps I should follow to use your code for my custom dataset?

Thank you!

bibekyess commented 1 year ago

Hi @DesaleF! Did you figure out how to do it? I can retrain the custom dataset from scratch but did you figure out how to fine tune it on the top of pretrained model (on VG dataset)?

bibekyess commented 1 year ago

Maybe this helps: #180

majianbo3 commented 1 month ago

Maybe this helps: #180

Hello, I have been learning about scene maps recently, but I am not sure how to create my own training set. If you know how to create it and let me know, I would be very grateful

zzzcccxx commented 3 weeks ago

Maybe this helps: #180

Hello, I have been learning about scene maps recently, but I am not sure how to create my own training set. If you know how to create it and let me know, I would be very grateful

Hello, do you now know how to create dataset?