search

HRNet / HRNet-Semantic-Segmentation

The OCR approach is rephrased as Segmentation Transformer: https://arxiv.org/abs/1909.11065. This is an official implementation of semantic segmentation for HRNet. https://arxiv.org/abs/1908.07919
Other
3.13k stars 686 forks source link

performance of pascal-context #240

Open Margrate opened 3 years ago

Margrate commented 3 years ago

I got mIoU=51.42 (59 classes) and mIoU=46.58(60 classes) which was a great gap compared with paper image

johnny-wang16 commented 3 years ago

Hi there I got mIOU also around 51 for 59 classes. But it took me at least 2 weeks to train on 4 GPUs (TiTan Xp) for 200 epochs. May I ask whats your training time ?

Margrate commented 3 years ago

You should test with flip and multi scale. My training time is about 20 hours

---Original--- From: @.> Date: Sun, Sep 12, 2021 20:57 PM To: @.>; Cc: @.**@.>; Subject: Re: [HRNet/HRNet-Semantic-Segmentation] performance of pascal-context (#240)

Hi there I got mIOU also around 51 for 59 classes. But it took me at least 2 weeks to train on 4 GPUs (TiTan Xp). May I ask whats your training time ?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android.

Margrate commented 3 years ago

加我的微信交流  1137096314

---Original--- From: @.> Date: Sun, Sep 12, 2021 20:57 PM To: @.>; Cc: @.**@.>; Subject: Re: [HRNet/HRNet-Semantic-Segmentation] performance of pascal-context (#240)

Hi there I got mIOU also around 51 for 59 classes. But it took me at least 2 weeks to train on 4 GPUs (TiTan Xp). May I ask whats your training time ?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android.

johnny-wang16 commented 3 years ago

Hi Margrate. Thanks for the info! which branch are you on and do you have to change any code to get it working ?