Comments (4)
Hi, thank you for your question. We did not set the specific dimension for "F.softmax" at Line 70 in /align/get_nets.py, which has been deprecated. You can optionally include a "dim=X" ("X" is the output neuron number) to avoid this warning. For the second warning, you can remove ", volatile = True" from "img_boxes = Variable(torch.FloatTensor(img_boxes), volatile = True)" at Line 82 in /align/detector.py and put the relevant code within "with torch.no_grad():". Hope this helps:)
from face.evolve.
@ZhaoJ9014 Thanks for your dedicate reply. BTW, how much MTCNN fps runs on CPU for now? Haven't test on a video but image seems fast. And is there any training process included to train a MTCNN face detector?
from face.evolve.
Welcome. 1) We haven't performed video test on CPU, but the speed should be also satisfactory. 2) Currently, we did not include the training process due to a common issue that the re-implementations can hardly reproduce the results reported by the MTCNN original paper. We thus choose to transfer the weights and implement the inference process. Please kindly inform us if you get any new results by performing video test on CPU or reproducing the reported results with self-training in the future.
from face.evolve.
Hi, thank you for your question. We did not set the specific dimension for "F.softmax" at Line 70 in /align/get_nets.py, which has been deprecated. You can optionally include a "dim=X" ("X" is the output neuron number) to avoid this warning.
specifically in this case, how many output neuron number of this model.
I using Onet, Pnet and Rnet
Btw, i don't see any flatten layer or linear layer in Onet, Pnet or Rnet
Thanks you
from face.evolve.
Related Issues (20)
- convert model to tensorflow lite
- 戴口罩的人脸识别有做嘛?
- Could you please help with pretrained model of IR_SE_50?
- CFP aligned dataset HOT 1
- how to do inference on webcam? HOT 1
- How to implement the triple loss? HOT 1
- Private Asia Face Data
- paddle inference demo 无法运行 HOT 2
- quero ve fecha essa po
- ms1mv2
- Missing HEAD checkpoint file
- pretrained model seems wrong
- Can we use the pretrained models for commercial use?
- Slow training HOT 1
- does it support facial recognition with masks?
- How to make the Private Asia Face Data ?
- Accuracy, training loss is so bad
- how to set the margin of the face cropped after face alignment?
- 是不是粘贴错了什么东西(
- splitting the datasets
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from face.evolve.