Project to Analysis images using different computer vision Algorithm. to able to get all the info from an image. And save them to a DB Article: Computer vision projects
- Object Detection
- Deep Labelling for Semantic Image Segmentation
- Image Captioning
- Face Clustering
- Face Recognition
- OCR
- Detect image properties
- Detect Labels
- Landmark Detection
- Pose Estimation
- Reverse Image search
A micro service system, each service will do a specific task. we are using:
- GRPC
- docker
- TFX
- Python
- elastic
Save image and meta data in mongo DB
Send Image via grpc https://stackoverflow.com/questions/62171037/grpc-python-sending-image-meta-data
- Multi-task learning
- Different Pre-trained models
- TFX
- Apache Beam
- Apache Airflow
- KubeFlow
All pertained model should be downloaded and unzipped in the pretrained_model folder
- Mutli Task : https://www.ijcai.org/Proceedings/2018/0168.pdf (paper), https://github.com/andyweizhao/Multitask_Image_Captioning (code)
- https://www.tensorflow.org/tutorials/text/image_captioning
- Deepface
- Insightface (ArcFace with LResNet100E-IR)
https://github.com/tensorflow/models/tree/master/research/delf#delg
Using the object detection with Model trained on OID v4 data set
No dataset with name: Landmark are only ID details
based on Celebrity in places we chosen images of this celebrities, also image from social media to generate a small data set, To test the pertained model before using them in our system.