CS257 2006 ghLecture 6

download CS257 2006 ghLecture 6

of 33

Transcript of CS257 2006 ghLecture 6

  • 7/29/2019 CS257 2006 ghLecture 6

    1/33

    CS257 Modelling Multimedia Information

    LECTURE 6

  • 7/29/2019 CS257 2006 ghLecture 6

    2/33

    Introduction

    See beginning of Lecture 5

  • 7/29/2019 CS257 2006 ghLecture 6

    3/33

  • 7/29/2019 CS257 2006 ghLecture 6

    4/33

    Queries to Video Databases

    User may want to specify a temporal

    sequence of events, e.g. find me video

    where this happens then this happens

    while that happens

    [More on this in PART 2]

  • 7/29/2019 CS257 2006 ghLecture 6

    5/33

    Queries to Video Databases

    How to express queries / How to describe

    content can be considered two sides of

    the same coin; both require dealing with

    the same kinds of issues

  • 7/29/2019 CS257 2006 ghLecture 6

    6/33

    Creating Metadata for Video Data

    Content-descriptive metadata for video oftenneeds to be manually annotated

    However, in some cases the process can be

    automated (partially) by: Video segmentation Feature recognition, e.g. to detect faces, explosions, etc.

    Extracting keywords from time-aligned collateral texts,e.g. subtitles and audio description

  • 7/29/2019 CS257 2006 ghLecture 6

    7/33

    Overview of LECTURE 6

    PART 1:Need to be able to formally describe video content interms of objects and events in order to make a query to a videodatabase, e.g. specify who is doing what.

    Subrahmanians Video SQL

    PART 2: May wish to specify temporal and / or causalrelationships between events, e.g. X happens before Y, A causes Bto happenAllens temporal logic

    Roths system for video browsing by causal links

    LABBring coursework questions;

  • 7/29/2019 CS257 2006 ghLecture 6

    8/33

    PART 1:Querying Video Content

    Four kinds of retrieval according to Subrahmanian (1998)

    Segment Retrieval: find all video segments where an

    exchange of a briefcase took place at Johns house

    Object Retrieval: find all the people in the video

    sequence (v,s,e)

    Activity Retrieval: what was happening in the video

    sequence (v,s,e)

    Property-based Retrieval: find all segments where

    somebody is wearing a blue shirt

  • 7/29/2019 CS257 2006 ghLecture 6

    9/33

    Querying Video Content

    Subrahmanian (1998) proposes an extension toSQL in order to express a users informationneed when querying a video database Based on video functions

    Recall that SQL is a database query languagefor relational databases; queries expressed interms of:

    SELECT (which attributes)FROM (which table)

    WHERE (these conditions hold)

  • 7/29/2019 CS257 2006 ghLecture 6

    10/33

    Subrahmanians

    Video FunctionsFindVideoWithObject(o)

    FindVideoWithActivity(a)

    FindVideoWithActivityandProp(a,p,z)

    FindVideoWithObjectandProp(o,p,z)

  • 7/29/2019 CS257 2006 ghLecture 6

    11/33

    Subrahmanians

    Video Functions (continued)FindObjectsInVideo(v,s,e)

    FindActivitiesInVideo(v,s,e)

    FindActivitiesAndPropsInVideo(v,s,e)

    FindObjectsAndPropsInVideo(v,s,e)

  • 7/29/2019 CS257 2006 ghLecture 6

    12/33

    A Query Language for Video

    SELECT may contain

    Vid_Id : [s,e]

    FROM may contain

    video : WHERE condition allows statements like

    term IN func_call

    (term can be variable, object, activity or property valuefunc_callis a video function)

  • 7/29/2019 CS257 2006 ghLecture 6

    13/33

    EXAMPLE 1

    Find all video sequences from the library

    CrimeVidLib1 that contain Denis Dopeman

    SELECT vid : [s,e]

    FROM video : CrimeVidLib1

    WHERE

    (vid,s,e) IN FindVideoWithObjects(Denis Dopeman)

  • 7/29/2019 CS257 2006 ghLecture 6

    14/33

    EXAMPLE 2

    Find all video sequences from the library

    CrimeVidLib1 that show Jane Shady giving

    Denis Dopeman a suitcase

  • 7/29/2019 CS257 2006 ghLecture 6

    15/33

    EXAMPLE 2

    SELECT vid : [s,e]

    FROM video : CrimeVidLib1

    WHERE

    (vid,s,e) IN FindVideoWithObjects(Denis Dopeman) AND

    (vid,s,e) IN FindVideoWithObjects(Jane Shady) AND

    (vid,s,e) IN FindVideoWithActivityandProp(ExchangeObject, Item, Briefcase) AND

    (vid,s,e) IN FindVideoWithActivityandProp(ExchangeObject, Giver, Jane Shady) AND

    (vid,s,e) IN FindVideoWithActivityandProp(ExchangeObject, Receiver, Denis Dopeman)

  • 7/29/2019 CS257 2006 ghLecture 6

    16/33

    EXAMPLE 3

    Which people have been seen with Denis

    Dopeman in CrimeVidLib1

  • 7/29/2019 CS257 2006 ghLecture 6

    17/33

    EXAMPLE 3

    SELECT vid : [s,e], Object

    FROM video : CrimeVidLib1

    WHERE

    (vid,s,e) IN FindVideoWithObject(Denis Dopeman) AND

    Object IN FindObjectsInVideo(vid,s,e) AND

    Object = Denis Dopeman AND

    type of (Object, Person)

  • 7/29/2019 CS257 2006 ghLecture 6

    18/33

    Exercise 6-1

    Given a video database of old sports broadcasts, calledSportsVidLib, express the following users information needsusing the extended SQL as best as possible. You shouldcomment on how well the extended SQL is able to captureeach users information need and discuss alternative ways of

    expressing the information need more fully.

    Bob wants to see all the video sequences with Michael Owen kicking a ball

    Tom wants to see all the video sequences in which Vinnie Jones is

    tackling Paul Gascoigne

    Mary wants to see all the video sequences in which Roy Keane is arguingwith the referee, because Jose Reyes punched Gary Neville, while ThierryHenry scores a goal, and then Roy Keane is sent off.

  • 7/29/2019 CS257 2006 ghLecture 6

    19/33

    Bob wants to see all the video sequences

    with Michael Owen kicking a ball

  • 7/29/2019 CS257 2006 ghLecture 6

    20/33

    Tom wants to see all the video sequences in which

    Vinnie Jones is tackling Paul Gascoigne

  • 7/29/2019 CS257 2006 ghLecture 6

    21/33

    Mary wants to see all the video sequences inwhich Roy Keane is arguing with the referee,

    because Jose Reyes punched Gary Neville, while

    Thierry Henry scores a goal, and then Roy Keane

    is sent off.

  • 7/29/2019 CS257 2006 ghLecture 6

    22/33

    Think about

    What metadata would be required in

    order to execute these kinds of video

    query?

    How could this be stored and searched

    most efficiently?

  • 7/29/2019 CS257 2006 ghLecture 6

    23/33

    Part 2: Enriching Video Data

    Models and Queries More sophisticated queries to video databases

    can be supported by considering: Temporal relationships between video intervals

    Causal relationships between events

    Need to be able to describe temporalrelationships between intervals formallyandmake inferencesabout temporal sequences

  • 7/29/2019 CS257 2006 ghLecture 6

    24/33

    Temporal Relationships

    between Intervals Allens (1983) work on temporal logic is often discussed

    in the video database literature (and in other computingdisciplines)

    13 temporal relationships that describe the possible

    temporal relationships that can hold between temporalintervals (e.g. intervals or events in video) these canbe used to formulate video queries

    A transitivity table allows a system to infer the relationshipbetween A r C, if A r B and B r C are known (where r

    stands for one temporal relationship, and A, B, C areintervals)

    SEE MODULE WEB-PAGE FOR EXTRA NOTES ON THIS

    X equal Y = = XXXXX

  • 7/29/2019 CS257 2006 ghLecture 6

    25/33

    X equal Y = = XXXXX

    YYYYY

    X before Y < > XXXX YYYY

    X meets Y m mi XXXXYYYY

    X overlaps Y o oi XXXXX

    YYYYY

    X during Y d di XXX

    YYYYYYYYY

    X starts Y s si XXXXYYYYYYYY

    X finishes Y f fi XXXXX

    YYYYYYYYYY

  • 7/29/2019 CS257 2006 ghLecture 6

    26/33

    Temporal Relationships

    between Intervals

    Crucial aspect of Allens work is the transitivity

    table that enables inferences to be made about

    temporal sequences

    Inferences take the form:If A r B, and B r C, then r1, r2, r3 may hold

    between A and C

    For example:If A < B and B < C, then A < C

  • 7/29/2019 CS257 2006 ghLecture 6

    27/33

    Another Example

    If A contains B, and B < C then whatrelationships can hold between A and C?

    BBBBB ?CC? ?CCCC? ?CCCCC?AAAAAAAAAAAAA?CCCCC?

    ?CCCCC?

    Possibilities:A < C ; A overlaps C; A meets

    C; A contains C; A is finished by C

  • 7/29/2019 CS257 2006 ghLecture 6

    28/33

    Modelling the Relationships between

    Entities and Events in Film

    Some temporal relationships might beinterpreted as causal relationships

    Roth (1999) proposed the use of a semanticnetwork to represent the relationships between

    entities and events in a movie includingcausal relations

    The user can then browse between scenes ina movie, e.g. if they are watching the scene of

    an explosion, they may browse to the scene inwhich a bomb was planted, via the semanticnetwork (extra note on semantic network willbe on the module website).

  • 7/29/2019 CS257 2006 ghLecture 6

    29/33

  • 7/29/2019 CS257 2006 ghLecture 6

    30/33

    Organising and Querying Video

    Content Should consider

    Which aspects of the video are likely to be ofinterest to the users who access the video

    archive? How to store relevant information about the

    video efficiently?

    How to express and process queries?

    What scope of automatic content extraction?

  • 7/29/2019 CS257 2006 ghLecture 6

    31/33

    EXERCISE 6-2 For an video database application domain of your

    choosing write five video queries that use some of

    Allens 13 temporal relationships

    If event A is before (

  • 7/29/2019 CS257 2006 ghLecture 6

    32/33

    LECTURE 6:

    LEARNING OUTCOMES

    After the lecture, you should be able to:

    Express a users query to a video database

    using Subrahmanians VideoSQL and discuss

    the limitations of this formalism Explain how and why temporal and causal

    relationships between events are represented in

    metadata for video databases

  • 7/29/2019 CS257 2006 ghLecture 6

    33/33

    OPTIONAL READING

    Dunckley (2003), pages 38-39; 393-395.For details of the extended video SQL, see:

    Subrahmanian (1998). Principles of Multimedia Databases- pages 191-195. IN LIBRARY ARTICLE COLLECTION

    For temporal relationships:Allen (1983). J. F. Allen, Maintaining Knowledge About TemporalIntervals. Communications of the ACM26 (11), pp. 832-843.Especially Figure 2 for the 13 relationships and Figure 4 for the fulltransitivity table. [In Library on shelf]

    For causal relationships:

    Roth (1999). Volker Roth, Content-based retrieval from digital video.Image and Vision Computing17, pp. 531-540. [Available onlinethrough library eJournals]