search

tim-learn / SHOT

code released for our ICML 2020 paper "Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptation"
MIT License
437 stars 78 forks source link

About Officehome #23

Closed oldfemalepig closed 3 years ago

oldfemalepig commented 3 years ago

Thank you very much for your work. What I want to ask is that I used a certain enhancement method to increase the "Source only" to 62. I think this enhances the generalization ability of the model to a certain extent. But when using the target domain for migration, the final effect is actually lower than the 71.8 in the paper?

tim-learn commented 3 years ago

Thank you very much for your work. What I want to ask is that I used a certain enhancement method to increase the "Source only" to 62. I think this enhances the generalization ability of the model to a certain extent. But when using the target domain for migration, the final effect is actually lower than the 71.8 in the paper?

Thank you very much for your work. What I want to ask is that I used a certain enhancement method to increase the "Source only" to 62. I think this enhances the generalization ability of the model to a certain extent. But when using the target domain for migration, the final effect is actually lower than the 71.8 in the paper?

I think the adaptation performance does not strictly correlate with that of the "source-only" baseline, so this experiment you conducted is quite normal. Maybe you could adjust the trade-off parameters or something else to test the adaptation performance.