3203. Introduction to Artificial Intelligence

Perception

 

1. Introduction

Perception is the process of acquiring, interpreting, selecting, and organizing sensory information.

Perception presumes sensation, where various types of sensors each converts a certain type of simple signal into data of the system. To put the data together and to make sense out of them is the job of the perception mechanism.

Perception can be seen as a special type of categorization (or classification, pattern recognition) where the inputs are sensory data, and the outputs are categorical judgments and conceptual relations.

The difficulty of the task comes from the need of multiple levels of abstraction, where the relations among data items are many-to-many, uncertain, and changing over time.

Accurately speaking, we never "see things as they are", and perception process of an intelligent system is often (and should be) influenced by internal and external factors beside the signals themselves. Furthermore, perception is not a pure passive process driven by the input.

In AI, the study on perception is mostly focused on the reproduction of human perception, especially on the perception of aural and visual signals. However, this is not necessarily the case since the perception mechanism of a computer system does not have to be identical to that of a human being.

 

2. Hearing

Speech recognition is the front-end of a system that can perceive and understand spoken language, as used in voice command interface and speech-to-speech translation.

There are several approaches toward speech recognition:

The acoustic-phonetic approach postulates that there exist finite, distinctive phonetic units (phonemes) in spoken language and that these units are broadly characterized by a set of acoustic properties. Even though the acoustic properties of phonetic units are highly variable, both with speakers and with neighboring sounds, it is assumed in the acoustic-phonetic approach that the rules governing the variability are straightforward and can be readily learned.

The pattern-matching approach represent a speech-pattern in the form of a mathematical model. A direct comparison is made between the unknown speech (the speech to be recognized) with each possible pattern learned in the training stage in order to determine the identity of the unknown.

The artificial intelligence approach attempts to do speech recognition using various AI techniques, such as knowledge-based systems or neural networks.

A related topic is speech synthesis, that is, translation from text to speech. After the text analysis capabilities pre-process the text (digit sequences, abbreviations, etc.) the pronunciations of most ordinary words and proper names are decided by the dictionary-based methods. Finally there are methods responsible for post-processing (prosodic phrasing, word accentuation, sentence intonation) and the actual speech synthesis. Here is an on-line demo. A major remaining problem is naturalness, especially context and meaning related adjustments (emotion, stress, tone, ...). To fully solve this problem, it is probably necessary to fully understand the meaning of the message and the purpose of the speech.

Music perception and composition are also studied in AI. For example, there are music works produced by a computer program, and some of them are in the styles of various classical composers.

 

3. Vision

Vision begins with a large array of measurements of the light reflected from object surfaces onto the eye. Analysis then proceeds in multiple stages, with each producing increasingly more useful representations of information in the scene.

Computational vision studies often follow three primary stages:

  1. Early representations may capture information such as the location, contrast, and sharpness of significant intensity changes or edges in the image. Such changes correspond to physical features such as object boundaries, texture contours, and markings on object surfaces, shadow boundaries, and highlights. In the case of a dynamically changing scene, the early representations may also describe the direction and speed of movement of image intensity changes.
  2. Intermediate representations describe information about the three-dimensional (3-D) shape of object surfaces from the perspective of the viewer, such as the orientation of small surface regions or the distance to surface points from the eye. Such representations may also describe the motion of surface features in three dimensions.
  3. Higher-level representations of objects describe their 3-D shape, form, and orientation relative to a coordinate frame based on the objects or on a fixed location in the world. Tasks such as object recognition, object manipulation, and navigation may operate from the intermediate or higher-level representations of the 3-D layout of objects in the world.
For relatively simple pattern recognition problems, neural network is often used to directly map input into output via a learning process. In recent years, hierarchical learning methods have made remarkable progresses on various problems, such as CAPTCHA.

Vision is not a pure input process. Eye movement has important impact on human visual perception. An active vision system is one that is able to interact with its environment by altering its viewpoint rather than passively observing it, and by operating on sequences of images rather than on a single frame. Also, there is some study on using the eye-gaze of a computer user in the interface to aid the control of the application.

 

4. High-level perception

By "higher-level perception", we mean how the given input data is categorized. While in low-level perception, the processing is mostly "bottom-up", i.e., the output is more or less a function of the input, in higher-level perception there are many more factors involved.

"One of the most important properties of high-level perception is that it is extremely flexible. A given set of input data may be perceived in a number of different ways, depending on the context and the state of the perceiver. Due to this flexibility, it is a mistake to regard perception as a process that associates a fixed representation with a particular situation. Both contextual factors and top-down cognitive influences make the process far less rigid than this." [more on this topic].

Examples: