Human emotions
The first site I visited,
Affective Computing Home Page at Massachusetts Institute of Technology, contains information about the different research areas in affective computing: Recognizing patterns of affect in sensed responses
Understanding and modeling the user?s emotional experience
Synthesizing emotions in machines
Affective computing applications
Interactions with affective computers
There are also links for information about the
researcher group and the goal of affective computing. Rosalind Picard is the founder and director of the Affective Computing Research Group at MIT and she is one of the researchers involved. There is also a list of current projects and readings for those who are interested in learning more. There is also a What?s New in Affective Computing site that discusses the new developments in Affective Computing. The Group also includes a statement of purpose and discusses the steps that are taken during all experiments and research. This Mission site also contains information regarding the necessity of safeguarding each person?s privacy. It encourages good encryption and calls for a "hardware/software fire-wall around each system that effectively prevents outside agents from obtaining access to any information about the user?s emotional life without the user?s personal consent."The Research on Human Emotions site explains three theories of emotion: The James-Lang Theory, Cannon-Bard Theory, and Schachter-Singer Theory. Some of the research on human emotions takes place at the
Berkeley Psychophysiology Lab, the Naval Health Research Center in San Diego, and of course the MIT Media Laboratory. "How can emotions can be utilized by and produced in autonomous agents?" This is what The University of Birmingham?s Cognition and Affect Project is trying to find out. The Psychophysiology WWW Directory, Geneva Emotions Research Group and the Brandeis Emotions Page are other sites that contain more information on affect and emotion. Also listed on this page are two current research projects in human emotion: Emotion Recognition in an Actor and The Frustration Experiment.I feel this site is appropriate for any person who has had at least one psychology course. The language is simple to follow and the concepts are linked to areas that will help one to better understand the material presented. All of the links from this page have been presented in this report.
I agree that research in the area of human emotions is a good thing. In order for specialists to be able to effectively treat emotional disorders, they need to understand the underlying mechanisms of emotions. I had heard of
smart rooms and smart cars before reading this site, so I wasn?t totally surprised by the extent to which MIT and other research facilities have taken Affective Computing. For example, the Lexus will remember up to two drivers? preferences for seat and mirror positioning and music preferences. Think of what your house could remember about you.Anyone who is interested in computers and emotions would be interested in this believable and authoritative site.
This section of the Affective Computing site from MIT Media Laboratory describes the
Prototype Sensing System , four types of monitors which are used to detect changes in a person?s emotional state by monitoring key body functions which are controlled by the Sympathetic Nervous System. The Galvanic Skin Response (GSR) Sensor measures the skin conductance, as a function of sweat gland activity, between two electrodes. The sweatier a person gets, the higher the skin conductance. As the anxiety level of the person decreases, so does the conductance due to reabsorption. Operating room personnel use the GSR and the Blood Volume Pulse Sensor, which uses light to detect blood pressure in the extremities, to monitor their patients. The Respiration Sensor is a large Velcro belt that is placed around the chest and it monitors the diaphragm by recording the amount the belt stretches. Runners use a respiration sensor to maximize their work outs. The Electromyogram Sensor measures muscle activity and this can be extremely helpful in cases such as jaw clenching. These components make up the Sensing module. The final part of the Sensing Human Affect Response site lists current projects in Affect Recognition: Affective Jewelry, Conductor's Jacket, Digital Processing of Affective Signals, Expression Glasses, The Frustration Experiment, Sentic Mouse, and Smart SHELL.This section is very comprehendible. I am able to understand how the equipment works. I agree that this is an effective way to measure a person?s physiological responses to emotion, but I do not feel it is an appropriate way to create labels for emotions because everyone is different. One person?s response to anxiety may be the exact response of another person to joy. A baseline for each body function would have to be established before any conclusions could be drawn from the data. Self-report would also be necessary to confirm or deny a researcher?s conclusions. The only sure thing a researcher could say is that the heart rate elevated during a startle or the heart rate dropped during a startle. Neither proves that the patient was scared or anxious.
I learned that there are many projects that are current and ongoing in the area of Affective Recognition. They are listed above, so all you have to do is click on the topic of interest and you will be taken to the site. I now understand how the Galvanic Skin Response Sensor and the Blood Volume Pulse Sensor work. That was a mystery to me before I read this page. Once again, anyone who is interested in computers and emotions would be interested in this site. These experiments are believable and authoritative. Check them out for yourself.
Recognizing Patterns of Affect in Sensed Responses
This site is more abstract to me. One of the main problems I have with the site is that psychologists have been unable to define a model of emotions. Categorizing and classifying data has been inconsistent according to this site. The beginning of this site reminds us that there are four methods for gathering physiological data. A diagram of a Markov model for affective states follows. The attempt here seems to be to establish the Recognition module. The end of the site lists the current projects in affective recognition:
Frustration Detection, The Frustration Experiment, Photobook, and Four Eyes, and Smart SHELL.Affective Understanding: Modeling and Responding to the User?s Emotional Experience
In this area, the model for affective computing appears to come to life. The Affective Understanding Module is the next in logical sequence after the
Recognition module. This part of the "system" will be able to use the information gathered by the Sensing module and process it. By maintaining a model of the user?s emotions, it will be able to "generate appropriate responses to the user that incorporate the user?s emotional state, the user?s cognitive abilities, and the user?s environmental situation." There is an explanation as to how the features of the module will be able to "absorb information" and remember it. The Affective Understanding module will be able to "model the user?s current mood," "model the user?s emotional life," and "apply the user affect model." It will also "update the user affect model" and "build and maintain a user-editable taxonomy of user preferences." There will be a "two-way communication with the system?s recognition module and it will eventually be able to build and maintain a more complete model of the user?s behavior." Finally, the affective understanding module will "eventually model the user?s context," "provide a basis for the generation of the synthetic system affect," and "ensure confidentiality and security" of the user and their emotional life. This section explains all of the above in great detail.Anyone who is interested in learning more about how these Modules work will find this section valuable. I found it very confusing because I have a difficult time comprehending models of cognition, and this seems to be a very detailed model. The information is believable and I am just overwhelmed with the amount of research in the affective field that is in progress.
Synthesizing Emotions in Machines
This section discusses the synthesis model of emotion, or constructing emotion. It also tells who in the MIT Media Laboratory is working on this project and gives links to current research and developments at other institutions. This is the shortest section of this site that I have found and it is also the most vague and the most difficult for me to relate to.
Affective Computing Applications
This section discusses the applications for affective computing, i.e. now that we have discussed what is affective computing, we need to discuss how it can be used. This site says that if software can respond to the user with a greater sensitivity, then mission accomplished. In entertainment, the
Affective CD player is "trained to track the (emotional) states of the user" by keeping track of the music selections or "musical instruments that can sense the musician's mood." An affective tutor "is able to sense the level of interest or frustration of the pupil and incorporate this information to improve ts teaching." There is also an emotionally intelligent copy machine whose "interface design enables people to communicate and vent their frustrations that arise from using the machine. A photograph retriever "can find you pictures that are of a particular mood." Again, as in several other pages in this site, there is a list of the current projects, such as Affective Jewelry, Conductor's Jacket, Expression Glasses, and Smart Rooms.If all of this has seemed abstract so far, then, this is the first page where all of the different modules come together so that the reader can actually see what are the ways in which affective computing can really work. Which mediums can one use for affective computing? You will find some answers right here on this page.
This section talks about wearables that allow us to create systems. They are entire systems "that are carried by the user, from the CPU and hard drive, to the power supply and all input/output devices." The different input and output devices are listed and compared. And because these computing systems have to be on your person, they are very small and require batteries for supplying power. The page ends with the Research Projects in Affective Computing. Most of which you have probably been exposed to.
If we finally do create a computer with human like emotions, what do you think we will do with it? What benefits will affective computing provide to the average person?
Rosalind Picard wants these affective computing devices to help the individual in being aware of their emotional histories so that they can become more efficient and in control of their lives. It is meant to sense the physiological changes in a person so that they can either alter their perceptions of a particular situation or remove themselves from it. But, for every good intention there is also a bad one. There will be those who want to use the computer's ability to sense emotion for their own selfish purposes. For example, the polygraph test measures changes in the same bodily functions as the sensors for affective computing devices. The polygraph is not considered a reliable test by some critics, yet others swear by the results it produces. I thik we will use it for job interviews and security interviews. We may employ affective computing in the court room. We may put our presidents in an affective computing booth so that we can know for sure whether he is telling the truth or not.
How far has the research in this area gone?
Do you think there would be social controversy in dealing with affective computing, such as how can computers discriminate individual's unique emotion?
I believe there will be controversy surrounding affective computing and its devices once the progress has gone so far that the items are marketable to the public and businesses. Rosalind Picard favors very strong encryption of the emotional histories of the users so that they are the only ones who can access that information.
How will computers know the exact cause of the emotion, especially if it stems from multiple causes, in order to make suggestions about how to modify behavior?
These devices will hold the user's emotional history?that means that for years, the user programs the scenario's into the device and correlates emotions and responses to that affect with the setting. Remember that the system is only as good as it has been programmed.
The pros of affective computing - self awareness, self help, and better overall health for the user. The cons of affective computing - people may try to access users' histories in order to reject applicants for insurance or jobs. Another con is that the system only knows what you tell it, so if you forget to let the wearable or your house know that your mother just died, then the user could get really frustrated with that affective computing system.
I do not want a computer to know or recognize my anxiety in certain situations. I know that this could allow the computer to help me, but sometimes I am not ready to be told what someone or something else thinks about me. I also fear the abuses within the society. Insurance companies could pay hackers big money to access personal affective computing systems in order to gain more information about the risks of heart disease or panic attacks (neither which I have).
Smart rooms have cameras and other sensors that "try to interpret what people are doing in order to help them." They recognize who is in the room. They can interpret body language. The advantages for this for me would be that after a long hard day, I would walk into my home and my house would sense my mood in order to put on the most soothing music. The lights would turn on automatically in the rooms where I needed to work. If I had to prepare dinner, I would be able to know that my Smart House has already begun preheating the oven. There are also smart cars. They also have input devices such as cameras and microphones that know when drivers are going to make a turn, stop, or pass a car.
All of the sections of this site are very well organized. They begin with explanations of the topic. Whether it be definitions of terms or specific examples, the first section is always explanatory. The second part of these pages contains an in depth look into a particular aspect of affective computing. The final section of most of the pages listed above, ends with a list of current research projects that are specific to that topic of discussion. MIT has done a great job of organizing their pages with informative and interesting text. There are a good number of links within the document which allows the reader to switch back and forth between computer windows with ease. And if you should happen to lose your way, at the bottom of every page is a clickable navigation table.
In my answers to my classmates, I expressed some personal fears that stem from abuse of the technological advances. Hackers are an asset to the government and to corporations. So why wouldn't there be people who try to access others' emotional information in order to save themselves money. If a life insurance company knows that this person has a history of depression by secretly viewing his or her emotional history, then they would be able to deny them coverage on the off chance that this person might commit suicide. I mean, they wouldn't get their 50 years worth of premiums out of them before they die? Why wouldn't the government want the emotional history of a potential spy? They could prevent any possibility of espionage, if they know that this person is strong and loyal and not easily intimidated.
I do believe that affective computing has begun with positive goals and the ambitions of Rosalind Picard to help the people to have a better relationship with computers is admirable. She wants to find another way in which computers can help make our lives easier. She is interested in
Report Evaluation as Compared with My Generation
At this time, most members of Generation 11 have not yet posted their reports for me to compare mine against.
Another site that I visited is the
Digital Culture Site. This contains a nice summary of the same information contained in the Affective Computing site.My Home Page // My Generations File // My Newsgroups File // Affective Computing Oral Presentation // Life on the Screen: Chapter 8 Summary // My Report 1: Psychology of Computer Viruses // My Report 2: Psychology of Virtual Online Communities // Icons Folder // My Generation's Home Page Index // For AOL Users // Helpful Hints for Surviving // Virus Joke list // Index to All Generations // Home Page of Dr. Leon James // Send Me a Message // Go Back to the Top