Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
language:
|
| 3 |
+
- en
|
| 4 |
+
license: mit
|
| 5 |
+
library_name: keras
|
| 6 |
+
tags:
|
| 7 |
+
- human-activity-recognition
|
| 8 |
+
- pose-estimation
|
| 9 |
+
- mediapipe
|
| 10 |
+
- tensorflow
|
| 11 |
+
datasets:
|
| 12 |
+
- guillherms/human-activity-pose_v4
|
| 13 |
+
pipeline_tag: image-classification
|
| 14 |
+
pretty_name: Human Activity Recognition MLP
|
| 15 |
+
---
|
| 16 |
+
|
| 17 |
+
# 🧠 Human Activity Recognition MLP
|
| 18 |
+
|
| 19 |
+
This model classifies human activities (reading, waving, office work, etc.)
|
| 20 |
+
using pose landmarks extracted from **MediaPipe Pose**.
|
| 21 |
+
|
| 22 |
+
## 📊 Architecture
|
| 23 |
+
- Multi-Layer Perceptron (Dense NN)
|
| 24 |
+
- Input: 132 pose landmark features
|
| 25 |
+
- Output: 10 activity classes
|
| 26 |
+
|
| 27 |
+
## 🧩 Files
|
| 28 |
+
- `mlp_pose.h5`: Trained Keras model
|
| 29 |
+
- `label_encoder.pkl`: Encodes activity labels
|
| 30 |
+
|
| 31 |
+
## 🧠 Author
|
| 32 |
+
Created by **Guilherme Santos**
|