Controlling navigation
Control of navigation of a content item is carried out by detection of non-verbal sound events. On the basis of receipt of one or more non-verbal sound event reports, a navigation tool is provided with a corresponding sequence of navigation commands. The correspondence between navigation command sequences and non-verbal sound events is established through analysis or markup of the content item.
1. A computer for controlling navigation through a computer readable structured document, for production of an output to a user corresponding to that structured document, navigation being effected by one or more navigation commands, the computer comprising:
an audio receiver for receiving a stream of audio waveform samples;
a sound detector operable to detect on the stream of audio waveform samples, one or more non-verbal sound events, wherein the sound detector is operable to detect a sound event on the basis of one or more machine learned sound models, and, for each detected sound event to produce a sound identifier, each sound identifier comprising information identifying the corresponding sound event, and to construct a sequence of sound identifiers corresponding to detected sound events in the stream of audio waveform samples;
a sequence to sequence decoder operable to receive a sequence correspondence model for the structured document, the sequence correspondence model encoding correspondence between one or several sound event identifiers and one or several navigation commands for navigating through the structured document, the sequence to sequence decoder being further operable, on the basis of the sequence correspondence model, to convert the sequence of sound identifiers into a sequence of one or more navigation commands for navigating through the structured document;
a content navigator, responsive to navigation commands from the sequence to sequence decoder, to cause navigation through the structured document and generation of a corresponding output to a user.
2. A computer in accordance with claim 1 wherein the sequence to sequence decoder is operable to acquire the sequence correspondence model from the structured document.
3. A computer in accordance with claim 2 wherein the sequence to sequence decoder is operable to acquire the sequence correspondence model by virtue of processing the structured document to identify one or more elements of the structured document that is associated with a detectable sound event.
4. A computer in accordance with claim 3 wherein the sequence to sequence decoder is operable to acquire the sequence correspondence model by way of text analysis of the structured document.
5. A computer in accordance with claim 3 or claim 4 wherein the sequence to sequence decoder is operable to acquire the sequence correspondence model by way of image analysis of the structured document.
6. A computer in accordance with claim 1 , configured to receive a human input action, and wherein the sequence to sequence decoder is operable to acquire a sequence correspondence model on the basis of a received human input action.
7. A computer implemented method of controlling navigation through a structured document, for production of an output to a user corresponding to that structured document, navigation being effected by one or more navigation commands, the method comprising:
receiving a stream of audio waveform samples;
detecting, on the stream of audio waveform samples, one or more non-verbal sound events, on the basis of one or more machine learned sound models, and, for each detected sound event:
producing a sound identifier, each sound identifier comprising information identifying the corresponding sound event, and
constructing a sequence of sound identifiers corresponding to detected sound events in the stream of audio waveform samples;
receiving a sequence correspondence model for the structured document, the sequence correspondence model encoding correspondence between one or several sound event identifiers and one or several navigation commands for navigating through the structured document;
converting, on the basis of the sequence correspondence model, the sequence of sound identifiers into a sequence of one or more navigation commands for navigating through the structured document;
causing navigation through the structured document responsive to the sequence of one or more navigation commands, and generating a corresponding output to a user.
8. A computer implemented method in accordance with claim 7 wherein the converting comprises acquiring a sequence correspondence model from the structured document.
9. A computer implemented method in accordance with claim 8 wherein the sequence correspondence model is acquired by processing the structured document to identify one or more elements of the structured document that is associated with a detectable sound event.
10. A computer implemented method in accordance with claim 9 wherein the acquiring of the sequence correspondence model is by way of text analysis of the structured document.
11. A computer implemented method in accordance with claim 9 wherein the acquiring of the sequence correspondence model is by way of image analysis of the structured document.
12. A computer implemented method in accordance with claim 7 , wherein the converting comprises acquiring a sequence correspondence model based on a received human input action.