TensorFlow, Docker & GoLang - All for Image Rekognition Sangam Biradar(EngineITops)
•
1 like•431 views
presented : https://events.docker.com/events/details/docker-bangalore-presents-docker-k8s-iot-mldl-docker101-hands-on-workshop-for-beginners/ 21 sept 19 Docker, K8s, IoT & ML/DL + Docker101 Hands-On Workshop for Beginners
Report
Share
More Related Content
What's hot
What's hot (20)
Similar to TensorFlow, Docker & GoLang - All for Image Rekognition Sangam Biradar(EngineITops)
Similar to TensorFlow, Docker & GoLang - All for Image Rekognition Sangam Biradar(EngineITops) (20)
More from sangam biradar
More from sangam biradar (17)
Recently uploaded
Recently uploaded (20)
TensorFlow, Docker & GoLang - All for Image Rekognition Sangam Biradar(EngineITops)
- 1. TensorFlow + Golang + Docker = all for image recognition 21 Sept 2019 Docker Bangalore Meetup at Amazon Bangalore
- 2. $Whoami Sangam Biradar Github-sangam14 Twitter-BiradarSangam $curl engineitops.com - My Personal Blog Website - Over 100+ blog post on docker | Golang and kubernetes and cloud - Unique 50000+ hits - Research Papers Presented/Published in the Conference Proceedings(IEEE) 1.“Build Minimal Docker Container Using Golang’’- presented and published in the proceedings of the International Conference ICICCS2018 at Vaigai College Of Engineering, Madurai 14 June 18
- 3. Agenda - Introducing TensorFlow - TensorFlow with Golang - Implementation - Demo – - Questions?
- 4. Introduction to tensor flow Tensors, in general, are simply arrays of numbers, or functions, that transform according to certain rules under a change of coordinates. TensorFlow is an open source software library for doing graph- based computations quickly. It does this by utilizing the GPU(Graphics Processing Unit), and also making it easy to distribute the work across multiple GPUs and computers.
- 6. Common-use-cases you can find some in the TensorFlow Models repo https://github.com/tensorflow/models. we'll use one of them, called Inception to recognize an image. https://github.com/tensorflow/models/tre e/master/research/inception/inception
- 7. Tensor flow trained model image with TensorFlow plus Go to reduce Dockerfile. https://github.com/ctava/tensorflow-go Download Inception data: http://download.tensorflow.org/models/inception5h .zip Let's start with simple main.go file to test if our Dockerfile works.
- 8. Test reduce my Dockerfile. https://github.com/ctava/tensorflow-go Download Inception data: http://download.tensorflow.org/models/inception5h.zip Let's start with simple main.go file to test if our Dockerfile works. package main func main() { if len(os.Args) < 2 { log.Fatalf("usage: imgrecognition <image_url>") } fmt.Printf("url: %sn", os.Args[1]) }
- 9. Build Dockerfile docker build -t imgrecognition . docker run imgrecognition https://www.iaspaper.net/wp- content/uploads/2017/10/Rabbit-Essay.jpg Let's get our image from the provided URL: // Get image from URL response, e := http.Get(os.Args[1]) if e != nil { log.Fatalf("unable to get image from url: %v", e) } defer response.Body.Close()
- 10. Golang https://github.com/sangam14/Tenserflow-golang-docker-image- recongnition/blob/master/main.go Let's see: - NewTensor converts from a Go value to a Tensor - Build a graph of our image - Init a session, because all Graph operations in Tensorflow are done with sessions. - Run the session to normalize image, using input and output. - normalized[0] contains normalized Tensor. - In makeTransformImageGraph we define the rules of normalization.
- 11. Also let's skip those warnings: os.Setenv("TF_CPP_MIN_LOG_LEVEL", "2") Here we worked with pre-trained model, let's try this program with something unusual, like ... Gopher. docker run imgrecognition https://i.pinimg.com/736x/12/5c/e0/125ce0baff3271761ca61843eccf7985.jp g Mouse?? no! But it's possible to train our models from Go in TensorFlow.
- 12. Clone git repo. [node1] (local) root@192.168.0.8 ~ $ git clone https://github.com/sangam14/Tenserflow-golang- docker-image-recongnition Cloning into 'Tenserflow-golang-docker-image-recongnition'... remote: Enumerating objects: 15, done. remote: Counting objects: 100% (15/15), done. remote: Compressing objects: 100% (10/10), done. remote: Total 15 (delta 3), reused 12 (delta 3), pack-reused 0 Unpacking objects: 100% (15/15), done.
- 13. Clone git repo. $ cd Tenserflow-golang-docker-image-recongnition/ [node1] (local) root@192.168.0.8 ~/Tenserflow-golang-docker-image-recongnition [node1] (local) root@192.168.0.8 ~/Tenserflow-golang-docker-image-recongnition $ docker build -t imgrecognition . [node1] (local) root@192.168.0.8 ~/Tenserflow-golang-docker-image-recongnition $ docker run imgrecognition https://i.pinimg.com/736x/12/5c/e0/125ce0baff3271761ca61843eccf7985.jpg url: https://i.pinimg.com/736x/12/5c/e0/125ce0baff3271761ca61843eccf7985.jpg label: mouse, probability: 14.93% label: pick, probability: 10.40% label: wall clock, probability: 7.56% label: shield, probability: 5.54% label: hook, probability: 4.72%
- 14. Thanks Find demo: http://dockerlabs.collabnix.com/presentation/ten serflow-go.html#/ Visit : http://engineitops.com/tenser-flow- golang-docker-all-for-image-recognition-at- amazon-india-bangalore-sangam-biradar/