RangRang - Machine Learning

Machine Learning side of RangRang

To Do's

1. Create data for object detection and color detection level ✔️
2. Labelling images ✔️
3. Create the model using tensorflow object detection API ✔️

Update on Object Detection Model

Fine tune ssd_mobilenet_v2_fpnlite_320x320_coco17_tpu-8 model with data on the workspace/object_detection/images

Update on Color Detection Model

Fine tune ssd_mobilenet_v2_fpnlite_320x320_coco17_tpu-8 model with data on the workspace/color_detection/images

Dataset

• Dwiki Kusuma : living room object data on Kaggle

Scripts

scripts
├── add_metadata.py
├── check_image.py
├── convert_tflite.py
├── image_downloader.py
├── label_processing.py
├── preprocessing
│   ├── generate_tfrecord.py
│   ├── partition_dataset.py
│   └── renamer.py
└── query.json

References

1. Object Detection API with TensorFlow 2
2. Training Custom Object Detector
3. Transfer Learning Object detection model - TensorFlow 2 Detection Model Zoo
4. Running TF2 Detection API Models on mobile
5. Real-time Object Detection using SSD MobileNet V2 on Video Streams
6. Color Detection

Navigation

	Repository
Android	grrrracia/RangRang-MobileApp
Cloud Computing	Hyuto/rangrang-server
Machine Learning	Hyuto/rangrang-ML