First systematic contact "human-computer interaction" concept was at university, there is a professional course "Instructional Media and Technology", at present what they have learned nothing left, just remember this is the time when professional courses, very interested. Human-computer interaction - way of dealing with people and machines.
In the form of human-computer interaction upgrade three aspects of what probably happened:
1. human-computer interaction to mouse and keyboard for the media;
2. Touch screen, stylus, etc. as a medium of human-computer interaction;
3. voice as a carrier of human-computer interaction; the future may have more forms of interpersonal interaction.
In the process of development, in fact, the biggest change is the transition from the image to the interactive voice. Sound carries human thought and emotion, the image bearing human appearance and intuition. Visible graphical interaction more accessible to the user's attention, the first impression is determined by the appearance of humans. Although the larger the amount of information carried by the media image, however, the interactive graphical interface need to touch the trigger finger, over one arm free operating range of the far field of interactive, graphical interface might have barriers. Here, the voice interaction advantage within a certain range. It is also one of the greatest charm of voice interaction: hands free! Relative to the image interactive, voice interaction also has its advantages, but it still voice interaction problems.
Voice interaction, the main problems are:
1. Voice interaction wake unnatural voice interaction wake is currently issuing voice commands to wake up the machine, this wake-up mode is not natural. Under living life scenes, everyone will interact by voice, gestures, body movements, distance, lip movement mirrored interaction. To solve this problem can evoke evoke way to add more voice, to evoke a variety of ways to coexist voice interaction. Is determined by several dimensions to interact evoked, the first dimension may be determined from near field or far-field; second dimension gestures, body movements; third dimension is the lip-sound.
The robot no feelings, and too rigid robot interaction EQ consists of two parts, one emotion recognition, part of the emotional response. For emotion recognition, by locking face, image analysis methods to understand the user's mood, active trigger; method analysis of natural language, words and emotions mark analysis to identify the user's emotions. For emotional response, on the user sentiments reflected by analyzing the user's input and analyze several rounds of dialogue. Combine the previous emotion recognition, voice can be the basis of semantic understanding on identifying semantic speech here at the time of the occurrence, the user's emotion is how the situation.
3. The interactive behavior of the robot is not active, mostly passive interaction, dialogue inequality; 85% of users have to actively look forward to the exchange, the current robot human-computer interaction, mostly triggered by passive, not proactive, allowing users feel the feel of attention. At the same time, is also in the process of inequality mechanisms for dialogue, it will interact with the user after only when the machine is triggered, while not actively produce recognize people's emotions, called "the wind blows." The main problem with such unequal interaction is also currently exist.
Currently mainly used in human-computer interaction: chat, Q & A, the dialogue of these three scenarios.
Chat: robot evoke passive Q & A: be represented by a question and answer form, there is a lack of depth in context of answering system. Most expert systems to render certain areas of expertise.
Demo site: github.crmeb.net/u/LXT