ambiser
/
human
mirror of https://github.com/vladmandic/human
9
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

added experimental movenet-lightning

master
Vladimir Mandic 2021-05-29 09:18:00 -04:00
parent 30aee32aa1
commit 317a8fc76c
2 changed files with 11 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
Configuration.md

@ -133,10 +133,10 @@ const config: Config = {
body: {
enabled: true,
modelPath: 'posenet.json', // body model, can be absolute path or relative to modelBasePath
// can be 'posenet' or 'blazepose'
// can be 'posenet', 'blazepose', 'efficientpose', 'movenet'
maxDetected: 1, // maximum number of people detected in the input
// should be set to the minimum number for performance
// only valid for posenet as blazepose only detects single pose
// only valid for posenet as other models detects single pose
minConfidence: 0.1, // threshold for discarding a prediction
},

10
Models.md

@ -15,10 +15,14 @@ Default models in Human library are:
## Notes
**Body detection** can be switched from `PoseNet` to `BlazePose` depending on the use case:
**Body detection** can be switched from `PoseNet` to `BlazePose`, `EfficientPose` or `MoveNet` depending on the use case:
- `PoseNet`: Works with multiple people in frame, works with only partial people
Best described as works-anywhere, but not with great precision
- `MoveNet`: Works with single person in frame, works with only partial people
Modernized and optimized version of PoseNet with different model architecture
- `EfficientPose`: Works with single person in frame, works with only partial people
Experimental model that shows future promise but is not ready for wide spread usage due to performance
- `BlazePose`: Works with single person in frame and that person should be fully visibile
But if conditions are met, it returns far more details (39 vs 17 keypoints) and is far more accurate
Furthermore, it returns 3D approximation of each point instead of 2D
@ -55,6 +59,8 @@ Default models in Human library are:
| NanoDet | 255K | nanodet.json | 7.3M | nanodet.bin | 229 |
| MB3-CenterNet | 197K | nanodet.json | 1.9M | nanodet.bin | 267 |
| FaceRes | 70K | faceres.json | 6.7M | faceres.bin | 524 |
| MoveNet-Lightning | 158K | movenet-lightning.json | 4.5M | movenet-lightning.bin | 178 |
| MoveNet-Thunder | 158K | movenet-thunder.json | 12M | movenet-thunder.bin | 178 |
<br>
@ -71,6 +77,8 @@ Default models in Human library are:
- Hand Detection & Skeleton: [**MediaPipe HandPose**](https://drive.google.com/file/d/1sv4sSb9BSNVZhLzxXJ0jBv9DqD-4jnAz/view)
- Body Pose Detection: [**BlazePose**](https://drive.google.com/file/d/10IU-DRP2ioSNjKFdiGbmmQX81xAYj88s/view)
- Body Pose Detection: [**PoseNet**](https://medium.com/tensorflow/real-time-human-pose-estimation-in-the-browser-with-tensorflow-js-7dd0bc881cd5)
- Body Pose Detection: [**MoveNet**](https://blog.tensorflow.org/2021/05/next-generation-pose-detection-with-movenet-and-tensorflowjs.html)
- Body Pose Detection: [**EfficientPose**](https://github.com/daniegr/EfficientPose)
- Age & Gender Prediction: [**SSR-Net**](https://github.com/shamangary/SSR-Net)
- Emotion Prediction: [**Oarriaga**](https://github.com/oarriaga/face_classification)
- Face Embedding: [**BecauseofAI MobileFace**](https://github.com/becauseofAI/MobileFace)