Deep Learning based on Computer Vision
2 likes90 views
The document discusses deep learning and computer vision technologies used for teaching computers to recognize objects such as cats. It highlights various computational tools and methods for training, including GPUs, FPGAs, and Azure's capabilities for large-scale AI training and experimentation. Additionally, it outlines key features of the Azure Batch AI training service, emphasizing its flexibility and support for multiple frameworks.
![[8]](https://image.slidesharecdn.com/deeplearning-180906145121/85/Deep-Learning-based-on-Computer-Vision-20-320.jpg)
![[1,2,3]](https://image.slidesharecdn.com/deeplearning-180906145121/85/Deep-Learning-based-on-Computer-Vision-21-320.jpg)
![[1,2,3]
[4,5,6]
[7,8,9]](https://image.slidesharecdn.com/deeplearning-180906145121/85/Deep-Learning-based-on-Computer-Vision-22-320.jpg)
![[?]](https://image.slidesharecdn.com/deeplearning-180906145121/85/Deep-Learning-based-on-Computer-Vision-23-320.jpg)
Deep Learning based on Computer Vision
- 3. Agenda
- 8. Eu quero ensiar para o meu computador que isso é um gato O que é um gato? Eu sou um gato
- 9. Vamos ensinar esse computar
- 10. Isso é um gato. Agora eu já sei.
- 11. WTF?!?
- 15. MNIST
- 20. [8]
- 21. [1,2,3]
- 23. [?]
- 29. FPGAs EVALUATION CPUs and FPGAs, ASICs under investigation EFFICIENCY TRAINING CPUs and GPUs, limited FPGAs, ASICs under investigation Control Unit (CU) Registers Arithmetic Logic Unit (ALU) + + + + + + + FLEXIBILITY CPUs GPUs ASICs
- 32. Perceptron
- 34. What?
- 36. Nc FP32/64 TRAINING A/D SERIES AZURE BATCH ND2 INT32/64 TRAINING ND1 INT32 TRAINING INFINIBAND HIGH SPEED NETWORK FPGA1 ACCELERATED APPS FPGA MICROSERVICES HAAS MGMT Rendering Service HPC Azure AI Training Service • What is it? • Easy to run parallel large-scale training with GPU and InfiniBand high speed networking • Experimentation and simulation with any framework • Built on Azure Batch, Docker & Python • Azure Compliant Infrastructure • Designed with and supported by AIR Philly Team • • Key Capabilities • Open- TensorFlow, CNTK, Caffe, MXNet support • Submit and go parameterized job configuration • Experiment management and monitoring • Support teams with sharing of reserved capacity • Re-image VMs between users for compliance • Python API to integrate with experimentation tools • Publish models for eval with AML • H1 CY 2017 • Preview mid semester, summer GA –In Preview • Good support for internal and external early adopters AZURE BATCH DEEP LEARNING JOB SERVICE
- 37. Azure Batch AI Training AI Training At Scale with Azure