2018 03-07 Bright Night - Artificial Intelligence (AI)
Download as PPTX, PDF
1 like76 views
The document discusses various topics related to artificial intelligence including: - ImageNet, a database of images used for image recognition research. - How neural networks are trained using techniques like backpropagation and how network depth has improved accuracy. - Microsoft's cognitive services like computer vision, emotion recognition, and language understanding. - Research projects at Microsoft including generating artwork, machine translation, and modeling human interactions. - Standards for automated vehicle levels from SAE International. - Microsoft's principles for ethical AI design.
![Available
https://azure.microsoft.com/en-us/services/cognitive-services/computer-vision/
FEATURE NAME: VALUE
Description { "tags": [ "outdoor", "person", "building", "holding", "woman", "sidewalk",
"street", "young", "bicycle", "lady", "table", "walking", "red", "city", "girl",
"standing", "carrying", "man", "little", "wearing", "riding", "park", "phone",
"board", "kite", "hydrant" ], "captions": [ { "text": "Kelly Macdonald is
walking down the street", "confidence": 0.964099 } ] }
Tags [ { "name": "outdoor", "confidence": 0.9928818 }, { "name": "person",
"confidence": 0.9861265 }, { "name": "ground", "confidence": 0.9531361 }, {
"name": "sidewalk", "confidence": 0.7103443 }, { "name": "way",
"confidence": 0.576331437 } ]
Image format "Png"
Image dimensions 922 x 1382
Clip art type 0
Line drawing type 0
Black and white false
Adult content false
Adult score 0.0148117449
Racy false
Racy score 0.0157827977
Categories [ { "name": "people_", "score": 0.921875 } ]
Faces [ { "age": 28, "gender": "Female", "faceRectangle": { "top": 219, "left": 700,
"width": 115, "height": 115 } } ]
Dominant color background "White"
Dominant color foreground "White"
Accent Color #BF0C22](https://image.slidesharecdn.com/2018-03-07brightnight-ai-180308125413/85/2018-03-07-Bright-Night-Artificial-Intelligence-AI-11-320.jpg)
2018 03-07 Bright Night - Artificial Intelligence (AI)
- 4. X X X O X O X X O X O X O X X O X X O X X O X X O X X O X X O O X X O O X X O X O X O X O X O X O X O X O X O O X X X O O X X X O O X X X O O X X X O O X X
- 5. http://image-net.org/index ImageNet is an ongoing research effort to provide researchers around the world an easily accessible image database. The ImageNet project is inspired by a growing sentiment in the image and vision research field the need for more data. Ever since the birth of the digital era and the availability of web-scale data exchanges, researchers in these fields have been working hard to design more and more sophisticated algorithms to index, retrieve, organize and annotate multimedia data.
- 6. Input Output
- 10. We train neural networks with depth of over 150 layers. We propose a "deep residual learning" framework that eases the optimization and convergence of extremely deep networks. Our "deep residual nets" enjoy accuracy gains when the networks are substantially deeper than those used previously. Such accuracy gains are not witnessed for many common networks when going deeper. 028% 026% 016% 012% 007% 007% 004% 2010 2011 2012 AlexNet 2013 2014 VGG 2014 GoogleNet 2015 MS ResNet Human parity (5,1%) http://image-net.org/challenges/LSVRC/2015/results | https://blogs.microsoft.com/ai/microsoft-researchers-win-imagenet-computer-vision-challenge/ shallow 8 layers 19 layers 22 layers 152 layers
- 11. Available https://azure.microsoft.com/en-us/services/cognitive-services/computer-vision/ FEATURE NAME: VALUE Description { "tags": [ "outdoor", "person", "building", "holding", "woman", "sidewalk", "street", "young", "bicycle", "lady", "table", "walking", "red", "city", "girl", "standing", "carrying", "man", "little", "wearing", "riding", "park", "phone", "board", "kite", "hydrant" ], "captions": [ { "text": "Kelly Macdonald is walking down the street", "confidence": 0.964099 } ] } Tags [ { "name": "outdoor", "confidence": 0.9928818 }, { "name": "person", "confidence": 0.9861265 }, { "name": "ground", "confidence": 0.9531361 }, { "name": "sidewalk", "confidence": 0.7103443 }, { "name": "way", "confidence": 0.576331437 } ] Image format "Png" Image dimensions 922 x 1382 Clip art type 0 Line drawing type 0 Black and white false Adult content false Adult score 0.0148117449 Racy false Racy score 0.0157827977 Categories [ { "name": "people_", "score": 0.921875 } ] Faces [ { "age": 28, "gender": "Female", "faceRectangle": { "top": 219, "left": 700, "width": 115, "height": 115 } } ] Dominant color background "White" Dominant color foreground "White" Accent Color #BF0C22
- 16. Microsoft Research project https://blogs.microsoft.com/ai/drawing-ai/ The pictures are created by the computer, pixel by pixel, from scratch. This birds may not exist in the real world they are just an aspect of our computers imagination of birds. The drawing bot closes a research circle around the intersection of computer vision and natural language processing
- 17. Available https://azure.microsoft.com/en-us/services/cognitive-services/language-understanding-intelligent-service/ A machine learning-based service to build natural language understanding into apps, bots, and IoT devices. Quickly create enterprise-ready, custom models that continuously improve. Intents An intent represents actions the user wants to perform. Utterances An utterance is text input from the user that your app needs to understand. Entities An entity represents detailed information that is relevant in the utterance. Book me a flight to Amsterdam { query: Book me a flight to Amsterdam, topScoringIntent: { intent: BookFlight, score: 0.9887482 }, intents: [ { intent: BookFlight, score: 0.9887482 }, {
- 18. Available Break down language barriers with your friends, family and colleagues. Our online translator can help you communicate more clearly. Our voice translator currently works in 8 languages, Arabic, Arabic (Levantine), Chinese (Mandarin), English, French, German, Hindi, Italian, Japanese, Portuguese (Brazil), Russian, Spanish. and our text translator is available in more than 50 languages for instant messaging. Skype Translator uses machine learning. So the more you use it, the better it gets. https://www.skype.com/en/features/skype-translator/
- 21. Microsoft Research project The situated interaction research effort aims to enable computers to reason more deeply about their surroundings, and engage in fluid interaction with humans in physically situated settings. When people interact with each other, they engage in a rich, highly coordinated, mixed-initiative process, regulated through both verbal and non-verbal channels. In contrast, while their perceptual abilities are improving, computers are still unaware of their physical surroundings and of the physics of human interaction. Current human-computer interaction models assume there is a single user, engaging with full attention on a single computer system. Computers do not yet understand engagement, attention, proximity, interruptability, turn-taking, group dynamics, social expectations, human memory and goals, and so on. Our research aims to address these challenges and create the basis for a new generation of situated systems that are capable of fluid interactions and collaborations with people. conversation intention signal channel why: goals and intentions sense and reason about beliefs, intentions, goals, and long-term plans what: situation and activity sense and reason about relevant events and activities of self and others who: physical awareness Identify, track, and characterize relevant actors, objects, states, and relationships https://www.microsoft.com/en-us/research/project/situated-interaction/
- 22. Microsoft Research project https://www.microsoft.com/en-us/research/project/safe-autonomous-flight-everywhere-safe/ Microsoft open-sourced a research project called AirSim, a high-fidelity system for testing the safety of artificial intelligence systems. AirSim provides realistic environments, vehicle dynamics and sensing for research into how autonomous vehicles that use AI that can operate safely in the open world. AirSim has been developed as a plugin for Unreal Engine, a popular tool for game development. This means that the car simulation is decoupled from the environment it runs in. You can create an environment for your specific needs, such as a city or rural road, or choose from a variety of environments available online, and then simply drop in the AirSim plugin to test your self-driving algorithms in that environment. AirSim extensibility also allows researchers and developers to incorporate new sensors, vehicles or even use different physics engines.
- 24. SAE International is a global association of more than 128,000 engineers and related technical experts in the aerospace, automotive and commercial-vehicle industries. Available here: https://www.sae.org/standards/content/j3016_201401/ Level 0: No Automation Level 1: Driver Assistance Level 2: Partial Automation Level 3: Conditional Automation Level 4: High Automation Level 5: Full Automation The full time performance by an automated driving system of all aspects of the dynamic driving task under all roadway and environmental conditions that can be managed by a human driver The driving mode-specific performance of an automated driving system of all aspects of the dynamic driving task, even if a human driver does not respond appropriately to a request to intervene. The driving mode-specific performance by an automated driving system of all aspects of the dynamic driving task with the expectation that the human driver will respond appropriately to a request to intervene. The driving mode-specific execution by one or more driver assistance systems of both steering and acceleration/deceleration using information about the driving environment and with the expectation that the human driver perform all remaining aspects of the dynamic driving task. The driving mode-specific execution by a driver assistance system of either steering or acceleration/deceleration using information about the driving environment and with the expectation that the human driver perform all remaining aspects of the dynamic driving task. The full time performance by an automated driving system of all aspects of the dynamic driving task under all roadway and environmental conditions that can be managed by a human driver Humandrivermonitorsthedriving environment Automateddrivingsystem monitorsthedrivingenvironment
- 26. Build in transparency Use intelligible machines to educate users about how technology recognizes and analyzes information. Encourage algorithmic accountability Create new technologies with the expected and the unexpected in mind. Design for Privacy Adopt protection that secures personal and organizational information in ways that earn trust. Maintain dignity Prioritize preserving cultural commitments and empowering diversity. Guard against bias Ensure proper, representative research so that the wrong heuristics arent used to discriminate Augment human capabilities Design technology that assists humanity and respects human autonomy. More information: http://microsoft.com/ai At Microsoft, we believe that ethics and design go hand in hand. AI technology should not only be transparent, secure, inclusive, and respectful, but also maintain the highest degree of privacy protection. These principles guide the design of our own products and services and are what we aspire to be debated and adopted more broadly across industries and society.
- 28. AI and its role in society https://news.microsoft.com/futurecomputed/
- 29. microsoft.com/ai
- 30. microsoft.com/ai Studio is a powerfully simple browser-based, visual drag-and- drop authoring environment where no coding is necessary. Go from idea to deployment in a matter of clicks.
- 31. Risk Analytics Fraud Prevention Customer Next Best Action Predictive Maintenance Operations Automation Fleets management Personalization Dynamic pricing Advanced retail planning