Text to Video Converter
Publication Date: 2016-Jun-02
The IP.com Prior Art Database
System that takes an English language text and converts it into a video depicting the content of that text. This is valuable because some users may prefer to see a video instead of reading some text. There are many reasons why users may prefer video: some users may find video content easier to understand, or more interesting, or more entertaining. In some environments, video content can attract attention where text content may not. In some cases, users may understand content more thoroughly and/or remember it better if it is presented in video form. In addition, users with little or no reading ability may be able to understand a video even without being able to understand the text.
Page 01 of 7
Text to Video Converter
A system for producing video representation of a text, the system including, a text processing unit, a 3d modeling unit. The system operations include:
- accepting a text input
- utilizing the text processing unit to devise subject, object and relationships
- utilizing the 3d modeling unit to construct 3d objects associated with subject and object, animate said subject and object based on the relationship, render the animated object and subject into an output video
The 3d modeling unit including a 3d entity library. The 3d modeling unit associating subject and object devised by the text processing unit with 3d entities in the 3d entity library.
Two major phases: In the Design phase, system interprets the text and designs a video that depicts the content of that text. In the Implementation phase, system creates a video using the design generated in the previous phase.
The main advantages of this technology versus machine translation (or presenting the text in the original language) are : users may prefer video instead of text for a variety of reasons such as ease of understanding and degree of entertainment.
Working of system can be divided into two stages. Design and implementation
The Design Phase includes the following steps:
§ Initial Analysis
§ Break down the sentence into different Blocks
§ Compile a relationship diagram, which shows relation between the blocks . § Store required data in DB from diagram
The Implementation phase includes the following steps: § Create a physical model for each object.
§ Define the bones for the model in mesh.
§ Generate video using model and data from DB.
When an input text is entered, system uses a dictionary to interpret words in a language and convert them into concepts.
Consider the following example story:
"Once upon a time, a Hare and Tortoise lived in a forest. The Hare was very proud of his speed. "
· Count no. of words in the given input text and give numbering to all words.
No. of Words = 20
Page 02 of 7
1 2 3 4 5 6 7 8 9 101112 13 14 15 16 17 18 19 20
Once upon a time, a Hare and Tortoise lived in a forest. The Hare was very proud of his speed.
· Identify Parts of speech and abstract semantic types such as Living Being or Place or Inanimate Object.
· Remove known idioms that do not make any sense in video (employ a dictionary of idioms and check to see if any portion of the input matches).
Idiom : Once upon a time.
Noun : Living Beings: Hare, Tortoise
Verb : State Verb : lived, was
Adverb : Very
Pronoun : His
Adjective : Proud, speed
· Find the relationships between the words. There is extensive Enabling Art in the fields of Parsing, Semantic Relation Detection, and Semantic Role Labeling that involves finding relationships among entities. In the example sentence, a parser might find that "Hare" is one of the subjects of "lived".
· Use Coreference Resolution to determine whether nouns or pronouns in the same or different ent...