Recognition Task

A Recognition Task is a decisioning task that requires both detection and classification of input patterns (to identify and categorize objects, entityes, or patterns within data).

• AKA: Pattern Recognition Task, Recognition Problem, Identification and Classification Task.
• Context:
  • Task Input: Input Dataset, Pattern Type Set, Recognition Constraints
    • Optional Input: Recognition Metadata, Pattern Type Descriptions, Recognition Context
  • Task Output: Recognized Pattern, Recognition Category, Recognition Confidence Score
    • Optional Output: Recognizer Function, Recognition Features
  • Task Performance Measure: Recognition Accuracy, Recognition Precision, Recognition Recall, Recognition F-Measure
  • ...
  • It can typically decompose into Detection Tasks and Classification Tasks through recognition pipelines.
  • It can typically identify Pattern Instances via recognition algorithms.
  • It can typically assign Pattern Categorys using recognition models.
  • It can typically extract Pattern Features through recognition processing.
  • It can typically produce Structured Patterns from simple patterns.
  • It can typically measure Recognition Confidence via recognition scoring.
  • It can typically handle Recognition Ambiguity through recognition disambiguation.
  • ...
  • It can often support Resolution Tasks through recognition refinement.
  • It can often enable Normalization Tasks via recognition standardization.
  • It can often facilitate Extraction Tasks using recognition output.
  • It can often improve through Recognition Learning via recognition feedback.
  • It can often integrate Multiple Recognition Sources through recognition fusion.
  • ...
  • It can range from being a Simple Recognition Task to being a Complex Recognition Task, depending on its recognition pattern complexity.
  • It can range from being a Single-Modal Recognition Task to being a Multi-Modal Recognition Task, depending on its recognition input type.
  • It can range from being a Binary Recognition Task to being a Multi-Class Recognition Task, depending on its recognition category count.
  • It can range from being a Real-Time Recognition Task to being a Batch Recognition Task, depending on its recognition processing mode.
  • It can range from being a Supervised Recognition Task to being an Unsupervised Recognition Task, depending on its recognition training approach.
  • It can range from being a Perceptual Recognition Task to being a Cognitive Recognition Task, depending on its recognition abstraction level.
  • ...
  • It can be instantiated in Recognition Acts during recognition execution.
  • It can be solved by Recognition Systems implementing recognition algorithms.
  • It can utilize Recognition Models for recognition computation.
  • It can employ Recognition Frameworks for recognition organization.
  • ...
• Examples:
  • Language Recognition Tasks, such as:
    • Entity Recognition Tasks, such as:
      • Named Entity Recognition Task for person name identification.
      • Entity Mention Recognition Task for entity reference detection.
      • Product Named Entity Recognition Task for product name extraction.
    • Relation Recognition Tasks, such as:
      • Relation Mention Recognition Task for relationship identification.
      • Semantic Relation Recognition Task for meaning connection detection.
      • Protein Localization Mention Recognition Task for biological relation extraction.
    • Linguistic Pattern Recognition Tasks, such as:
      • Sentiment Recognition Task for emotion classification.
      • Part-of-Speech Recognition Task for grammatical category assignment.
      • Syntax Pattern Recognition Task for structure identification.
  • Visual Recognition Tasks, such as:
    • Object Recognition Tasks, such as:
      • Visual Entity Recognition Task for object identification.
      • Face Recognition Task for person identification.
      • Instance Recognition Task for specific object detection.
    • Pattern Recognition Tasks, such as:
      • Image Pattern Recognition Task for visual pattern detection.
      • Handwriting Recognition Task for text extraction.
      • Optical Character Recognition Task for printed text conversion.
    • Scene Recognition Tasks, such as:
      • Scene Understanding Task for environment interpretation.
      • Activity Recognition Task for action identification.
      • Gesture Recognition Task for movement interpretation.
  • Audio Recognition Tasks, such as:
    • Speech Recognition Tasks, such as:
      • Word Recognition Task for vocabulary identification.
      • Speaker Recognition Task for voice identification.
      • Language Identification Task for spoken language detection.
    • Sound Recognition Tasks, such as:
      • Music Recognition Task for song identification.
      • Environmental Sound Recognition Task for acoustic event detection.
      • Audio Pattern Recognition Task for sound pattern classification.
  • Biometric Recognition Tasks, such as:
    • Physical Recognition Tasks, such as:
      • Fingerprint Recognition Task for identity verification.
      • Iris Recognition Task for eye pattern matching.
      • Palm Print Recognition Task for hand identification.
    • Behavioral Recognition Tasks, such as:
      • Gait Recognition Task for walking pattern identification.
      • Keystroke Recognition Task for typing pattern detection.
      • Signature Recognition Task for handwriting verification.
  • Document Recognition Tasks, such as:
    • Text Recognition Tasks, such as:
      • Optical Character Recognition Task for printed text extraction.
      • Handwriting Recognition Task for cursive text conversion.
      • Historical Document Recognition Task for archival text processing.
    • Structure Recognition Tasks, such as:
      • Form Recognition Task for document field extraction.
      • Table Recognition Task for tabular data extraction.
      • Layout Recognition Task for document structure analysis.
  • Scientific Recognition Tasks, such as:
    • Molecular Recognition Tasks, such as:
      • Protein Structure Recognition Task for biological pattern identification.
      • Chemical Compound Recognition Task for molecular structure detection.
      • Gene Sequence Recognition Task for DNA pattern matching.
    • Signal Recognition Tasks, such as:
      • Radar Signal Recognition Task for target identification.
      • Medical Signal Recognition Task for diagnostic pattern detection.
      • Astronomical Signal Recognition Task for celestial object identification.
  • Security Recognition Tasks, such as:
    • Threat Recognition Tasks, such as:
      • Software Vulnerability Recognition Task for security flaw detection.
      • Intrusion Pattern Recognition Task for attack identification.
      • Fraud Pattern Recognition Task for suspicious activity detection.
    • Authentication Recognition Tasks, such as:
      • Person Recognition Task for identity verification.
      • Device Recognition Task for hardware identification.
      • Behavioral Pattern Recognition Task for user authentication.
  • ...
• Counter-Examples:
  • Detection Tasks, which only identify presence without classification.
  • Classification Tasks, which only categorize pre-detected objects without detection.
  • Segmentation Tasks, which partition data without necessarily recognizing patterns.
  • Disambiguation Tasks, which resolve ambiguity among known options rather than recognizing patterns.
  • Identification Tasks, which determine specific identity without requiring pattern classification.
  • Perception Tasks, which involve sensory processing without necessarily pattern recognition.
• See: Pattern Matching Task, Learning Task, Hypothesis Testing Task, Detection Algorithm, Recognition System, Concept Learning Task, Cognitive Task, Test Task.

---

---

References

2015

• (Wikipedia, 2015) ⇒ http://en.wikipedia.org/wiki/recognition#In_science_and_technology Retrieved:2015-4-3.
  • Automatic number plate recognition, the use of optical character recognition to read vehicle registration plates
  • Facial recognition system, a system to identify individuals by their facial characteristics
  • Iris recognition, a method of biometric identification
  • Antigen recognition, in immunology
  • Face perception, the process by which the human brain understands and interprets the face
  • Gesture recognition, the interpretation of human gestures
  • Handwriting recognition, the conversion of handwritten text into machine-encoded text
  • Intra-species recognition, the recognition of another member of the same species
  • Language identification, the problem of identifying which natural language given content is in
  • Magnetic ink character recognition, used mainly by the banking industry
  • Molecular recognition, the interaction between two or more molecules through non-covalent bonding
  • Named entity recognition, the classification of elements in text into predefined categories
  • Natural language understanding, the parsing of the meaning of text
  • Optical character recognition, the conversion of typewritten or printed text into machine-encoded text
  • Optical mark recognition, the capturing of human-marked data from document forms
  • Pareidolia, a psychological phenomenon in which a vague stimulus is perceived as significant
  • Pattern recognition, a branch of machine learning which encompasses the meanings below
  • Recall (memory), the retrieval of events or information from the past
  • Recognition (sociology), a public acknowledgement of person's status or merits
  • Recognition of human individuals, or biometrics, used as a form of identification and access control
  • Speech recognition, the conversion of spoken words into text

2009a

• (Wiktionary, 2009) ⇒ http://en.wiktionary.org/wiki/recognize
  • Verb to recognize (third-person singular simple present recognizes, present participle recognizing, simple past and past participle recognized)
    • 1. (transitive) To match something or someone which one currently perceives to a memory of some previous encounter with the same entity.
    • …

2009b

• (Wikipedia, 2009) ⇒ http://en.wikipedia.org/wiki/Recognition
  • Recognition is one of the three basic memory tasks. It involves identifying objects or events that have been encountered before.
  • Recognition (re+cognition) is a process that occurs in thinking when some event, process, pattern, or object recurs. Coming from the base cognition; cognition has various uses in different fields of study and has generally accepted to be used for the process of awareness or thought. In psychology, cognition is used for information processing view of a person's psychological functions. This takes place as we process the stimuli with previous memories and experiences and find relationships between the current stimuli and our memories.

2005

• (Strasburder, 2005) ⇒ Hans Strasburger. (2005). “Unfocussed Spatial Attention Underlies the Crowding Effect in Indirect Form Vision.” In: Journal of Vision, 5(11):8.
  • QUOTE: In a hierarchy of task complexity ranging from
    1. pattern detection (present/nonpresent),
    2. coarse grating discrimination1 (horizontal/vertical),
    3. fine grating discrimination (orientation threshold), and
    4. character recognition or identification,

Tasks 1 and 2 can be treated as more or less equivalent and as different from Task 4. The distinctive characteristic of the highest-level perceptual tasks–recognition or identification–as compared to the lower level (discrimination) we there suggested to be the dimensionality of the decision space, i.e. the requirement for the observer to chose his or her response from a comparatively large number of alternatives.

1The term “discrimination task” is sometimes used in a different meaning, implying the judgement of a quantity being larger or smaller than another (the corresponding psychometric function then goes from −1 to 1). This is not implied here, the intended meaning being that the observer can discriminate between two broadly different stimuli and thereby identify each. The term “identification task” is sometimes used for that case but is avoided here to reserve the concept of identification for those tasks where discrimination between a few cases will not solve the identification.

---