Eric Hu
- Total activity 50
- Last activity
- Member since
- Following 0 users
- Followed by 2 users
- Votes 0
- Subscriptions 20
Comments
Recent activity by Eric Hu Sort by recent activity-
Hi Char Stiles, Do you mind sharing your current frozen graph model before optimization? Also may I know what's current model size before optimization? Normally the optimization step only removes i...
-
Hi montek singh, Currently we only support converting models from ONNX or Tensorlfow's frozen graph (.pb) format. Thanks
-
Hi nolan_742, May I know did reducing the BATCH_SIZE help with the issue you found? Eric
-
Hi nolan_742, Yes just me is correct, everything looks fine and this issue is mainly because of Google Colab's GPU memory limit. To avoid this, we can do the following: 1. Reduce the BATCH_SIZE to ...
-
Hi ichabodcole, By clothing segmentation, I meant something like https://github.com/anish9/Fashion-AI-segmentation. Example training data can be found at https://github.com/bearpaw/clothing-co-pars...
-
Hi ichabodcole, Distinguishing neck from shoulder is definitely a hard segmentation problem due to vaguely defined border lines. Probably you can try segmenting clothes out and see if that can be a...
-
Hi j.patn, If you look at the Style Transfer notebook, under Global variables for training, you can change INPUT_HEIGHT and INPUT_WIDTH accordingly. Default is set as 512x256.
-
Hi Tom Geurts, Training a 6-DOF tracker involves many complications and we believe it's beyond the scope of our current ML tutorials. However, we will continue to provide support and may release mo...
-
Hi Irene, By default, our machine learning models should predict target segmentation regions as positive classes. In your case, we suspect that you are training your model with your region of inter...
-
Hi One More Admin, For now we are not planning to rewrite the same template with TensorFlow again mainly because we've found Pytorch is much easier for beginners to pick up and involves less variat...