What are the main courses in Artificial Intelligence?

Artificial intelligence technology is related to whether artificial intelligence products can be successfully applied to our life scenes. In the field of artificial intelligence, it generally contains seven key technologies: machine learning, knowledge graph, natural language processing, human-computer interaction, computer vision, biometric recognition, and AR/VR.

I. Machine Learning

Machine Learning (MachineLearning) is a cross-discipline involving statistics, system identification, approximation theory, neural networks, optimization theory, computer science, brain science, and many other fields, research on how computers can simulate or implement human learning behavior to acquire new knowledge or skills, and to reorganize existing knowledge structures to continuously improve themselves. Knowledge structure to continuously improve its own performance, is the core of artificial intelligence technology. Data-based machine learning is one of the most important methods in modern intelligence technology, which studies the search for laws from observed data (samples) and the use of these laws to make predictions about future data or unobservable data. Depending on the learning mode, learning method, and algorithm, there are different classifications of machine learning.

Machine learning is categorized as supervised learning, unsupervised learning, and reinforcement learning, etc. based on the learning mode.

Based on the learning method machine learning can be categorized into traditional machine learning and deep learning.

II. Knowledge Graph

Knowledge Graph is essentially a structured semantic knowledge base, a graph data structure consisting of nodes and edges, which describes concepts and their interrelationships in the physical world in a symbolic form, and whose basic unit of composition is the "Entity-Relationship-Entity Its basic unit is the "entity-relationship-entity" ternary, as well as entities and their related "attribute-value" pairs. Different entities are connected to each other through relationships, constituting a mesh-like knowledge structure. In the knowledge graph, each node represents an "entity" in the real world, and each edge is a "relationship" between entities. In layman's terms, the knowledge graph is a relational network obtained by connecting all different kinds of information together, providing the ability to analyze problems from the perspective of "relationships".

Knowledge graph can be used in anti-fraud, inconsistency verification, group fraud and other public **** safety and security field, need to use anomaly analysis, static analysis, dynamic analysis and other data mining methods. In particular, knowledge graphs have great advantages in search engine, visual display and precision marketing, and have become a popular tool in the industry. However, there are still great challenges in the development of knowledge graph, such as the problem of data noise, i.e., the data itself has errors or the data has redundancy. As the application of knowledge graph continues to deepen, there are a series of key technologies that need to be broken through.

Three, natural language processing

Natural language processing is an important direction in the field of computer science and artificial intelligence, the study of various theories and methods that can realize the effective communication between humans and computers in natural language, involving more fields, mainly including machine translation, machine reading comprehension and question and answer systems.

Machine translation

Machine translation technology refers to the use of computer technology to realize the translation process from one natural language to another natural language. Statistical-based machine translation methods break through the limitations of previous rule-based and instance-based translation methods, and the translation performance has been greatly improved. The successful application of deep neural network-based machine translation in some scenarios such as daily spoken language has shown great potential. With the development of contextual representation of the context and knowledge-based logical reasoning capabilities, and the continuous expansion of the natural language knowledge graph, machine translation will make more progress in areas such as multi-round dialog translation and chapter translation.

Semantic Understanding

Semantic understanding technology refers to the use of computer technology to realize the understanding of the text chapter, and to answer the questions related to the chapter. Semantic understanding focuses more on understanding the context and controlling the accuracy of the answers. With the release of the MCTest dataset, semantic understanding has received more attention and achieved rapid development, with the emergence of related datasets and corresponding neural network models. Semantic understanding technology will play an important role in intelligent customer service, product automated Q&A and other related fields, further improving the accuracy of Q&A and dialog systems.

Question and Answer Systems

Question and Answer systems are divided into open-domain dialog systems and domain-specific question and answer systems. Q&A system technology is the technology that allows computers to communicate with people in natural language like humans. People can submit questions expressed in natural language to a Q&A system, and the system will return answers with high relevance. Although Q&A systems have already seen a number of application products emerge, most of them are used in areas such as practical information service systems and smartphone assistants, and there are still problems and challenges in terms of Q&A system robustness.

Natural language processing faces four major challenges:

One is the existence of uncertainty at different levels, such as lexical, syntactic, semantic, pragmatic, and phonological;

Two is the unpredictability of unknown linguistic phenomena due to new vocabulary, terminology, semantics, and grammar;

Three is that the insufficiency of data resources makes it difficult to cover the complex linguistic phenomena;

Fourth, the ambiguity and intricate correlation of semantic knowledge is difficult to be described by simple mathematical models, and semantic computation requires nonlinear computation with huge parameters

Fourth, human-computer interaction

Human-computer interaction is mainly researched in the exchange of information between humans and computers, which mainly includes two parts of the exchange of information from humans to computers and from computers to humans, and it is an important peripheral technology in the field of artificial intelligence. Human-computer interaction is a comprehensive discipline closely related to cognitive psychology, ergonomics, multimedia technology, virtual reality technology and so on. The traditional information exchange between human and computer mainly relies on interaction devices, which mainly include input devices such as keyboards, mice, joysticks, data garments, eye trackers, position trackers, data gloves, pressure pens, and output devices such as printers, plotters, monitors, helmet-mounted displays, speakers, and so on. Human-computer interaction technologies include voice interaction, emotional interaction, somatic interaction, and brain-computer interaction, in addition to the traditional basic interaction and graphical interaction.

V. Computer Vision

Computer vision is the science of using computers to mimic the human visual system, allowing computers to have the ability to extract, process, understand, and analyze images and image sequences similar to humans. Automated driving, robotics, intelligent medicine and other fields all require computer vision technology to extract and process information from visual signals. Recently, with the development of deep learning, preprocessing, feature extraction and algorithmic processing are gradually integrated to form end-to-end artificial intelligence algorithmic technology. According to the problem solved, computer vision can be divided into five categories: computational imaging, image understanding, 3D vision, dynamic vision and video coding and decoding.

Currently, computer vision technology is developing rapidly and has a preliminary industrial scale. The future development of computer vision technology is mainly facing the following challenges:

One is how to better combine with other technologies in different application fields, computer vision can widely utilize big data in solving certain problems, which has gradually matured and can surpass human beings, while it cannot achieve high accuracy in certain problems;

The second is how to reduce the development time and labor cost of computer vision algorithms, currently, computer vision algorithms can not achieve high precision. time and labor costs, the current computer vision algorithms require a large amount of data and manual labeling, and require a long development cycle to achieve the required accuracy and time-consuming application areas;

Third, how to accelerate the design and development of new algorithms, with the emergence of new imaging hardware and artificial intelligence chips, the design and development of computer vision algorithms for different chips and data acquisition equipment is also one of the challenges. One of the challenges is the design and development of computer vision algorithms for different chips and data acquisition devices.

Six, biometric identification

Biometric identification technology refers to the individual physiological characteristics or behavioral characteristics of the individual identity identification and authentication technology. From the application process, biometric identification is usually divided into two stages: enrollment and identification. Enrollment phase through the sensor on the human body's biological representation of information collection, such as the use of image sensors on the fingerprints and faces and other optical information, microphone on the voice of the voice and other acoustic information collection, the use of data preprocessing and feature extraction technology to process the collected data, to get the corresponding features for storage.

The recognition process uses the same information acquisition method as the enrollment process to collect information, data preprocessing and feature extraction from the person to be recognized, and then compares and analyzes the extracted features with the stored features to complete the recognition. From the point of view of application tasks, biometric identification is generally divided into identification and confirmation of two tasks, identification refers to the process of determining the identity of the person to be identified from the repository, a one-to-many problem; confirmation refers to the process of comparing the information of the person to be identified to the information of a specific single person in the repository, and determining the identity of the person to be identified, a one-to-one problem.

Biometric identification technology involves a wide range of content, including fingerprints, palm prints, face, iris, finger veins, voiceprints, gait and other biometric features, and its recognition process involves a number of technologies such as image processing, computer vision, speech recognition, machine learning. At present, as an important intelligent authentication technology, biometric identification is widely used in finance, public **** security, education, transportation and other fields.

Seven, VR/AR

Virtual Reality (VR)/Augmented Reality (AR) is a new type of audio-visual technology centered on computers. Combined with related science and technology, within a certain range to generate and the real environment in the visual, auditory, tactile and other aspects of the highly similar digital environment. Users with the help of the necessary equipment and the digital environment of the object to interact, mutual influence, get close to the real environment feeling and experience, through the display equipment, tracking and positioning equipment, tactile sense of interaction equipment, data acquisition equipment, special chips and so on to realize.

Virtual reality/augmented reality can be divided into five aspects: acquisition and modeling technology, analysis and utilization technology, exchange and distribution technology, display and interaction technology, and technical standard and evaluation system according to different processing stages from the perspective of technical characteristics. Acquisition and modeling technology researches how to digitize and model the physical world or human creativity, and the difficulty is the digitization and modeling technology of the three-dimensional physical world; analysis and utilization technology focuses on analyzing, understanding, searching, and intellectualizing digital content, and the difficulty lies in the semantic representation and analysis of the content; exchange and distribution technology mainly emphasizes the large-scale digital content in a variety of network environments Circulation, conversion, integration and personalized services for different end-users, etc., the core of which is open content exchange and copyright management technology; display and exchange technology focuses on various display technologies and interaction methods that conform to human habitual digital content, with a view to improving people's cognitive ability to complex information, and the difficulty lies in the establishment of a natural and harmonious human-computer interaction environment; standard and evaluation system focuses on the research of virtual reality/ augmented reality basic resources, content cataloging and knowledge-based methods. Augmented reality basic resources, content cataloging, source coding and other normative standards and corresponding evaluation techniques.

Currently, the challenges faced by virtual reality/augmented reality are mainly reflected in four aspects, namely, intelligent acquisition, universal equipment, free interaction and perceptual integration. There are a series of scientific and technological problems in hardware platforms and devices, core chips and devices, software platforms and tools, and related standards and norms. In general, virtual reality/augmented reality presents the development trend of intelligent virtual reality system, seamless integration of virtual and real environment objects, and all-round and comfortable natural interaction

.