Classifying collision events using inertial and audio data

1. A handheld device comprising:

an inertial sensor to generate one or more motion signals;

a microphone to generate one or more audio signals;

a wireless transceiver to transmit information regarding a collision between the handheld device and an object; and

logic, implemented at least partly in one or more of configurable logic or fixed-functionality logic hardware, to:

obtain motion data associated with the one or more motion signals,

obtain audio data associated with the one or more audio signals,

detect the collision between the handheld device and the object based on the motion data and the audio data, and

classify the object as one of a person or a second handheld device based on filtered collision data, wherein classification of the object includes operations to: distinguish between a first audible sound generation quality of a first strike event and a different second audible sound generation quality of a second strike event, wherein the first strike event is of the handheld device contacting the second handheld device and the second strike event is of the handheld device contacting the person, wherein the first strike event is to have the first audible sound generation quality based at least in part on a material of the second handheld device, and wherein the second strike event is to have the different second audible sound generation quality based at least in part on a material of the person.

2. The handheld device of claim 1 , wherein the logic is to:

remove noise from the motion data and the audio data to obtain the filtered collision data.

3. The handheld device of claim 2 , wherein the logic is to:

identify one or more features of the collision, wherein the one or more features are to include one or more of a maximum value, a minimum value, a baseline value, a linear acceleration integration, a Euler angle integration, an audio frequency spectrum or a post-collision vibration frequency spectrum, and

compare the one or more features to training data to classify the object.

4. The handheld device of claim 2 , wherein the logic is to trigger one or more of a competitive score update or an audiovisual effect based on whether the object is classified as the person or the second handheld device.

5. The handheld device of claim 1 , further including one or more buffers, wherein the logic is to:

detect an availability of one or more new data samples, and

retrieve the motion data and the audio data from the one or more buffers in response to the availability.

6. The handheld device of claim 1 , wherein the logic is to:

determine, based on the motion data, that a difference between a peak acceleration and an average acceleration has reached a first threshold, and

determine, based on the audio data, that a difference between a peak audio level and an average audio level has reached a second threshold.

7. The handheld device of claim 1 , further including a housing having a form of a sword, saber, foil, epee, or blade.

8. An apparatus comprising:

logic, implemented at least partly in one or more of configurable logic or fixed-functionality logic hardware, to:

obtain motion data,

obtain audio data,

detect a collision between a handheld device and an object based on the motion data and the audio data, and

classify the object based on filtered collision data, wherein the object to be classified is one of a person or a second handheld device, and wherein the classification of the object includes: distinguishing between a first audible sound generation quality of a first strike event and a different second audible sound generation quality of a second strike event, wherein the first strike event is of the handheld device contacting the second handheld device and the second strike event is of the handheld device contacting the person, wherein the first strike event is to have the first audible sound generation quality based at least in part on a material of the second handheld device, and wherein the second strike event is to have the different second audible sound generation quality based at least in part on a material of the person.

9. The apparatus of claim 8 , wherein the logic is to:

remove noise from the motion data and the audio data to obtain the filtered collision data.

10. The apparatus of claim 9 , wherein the logic is to:

identify one or more features of the collision, wherein the one or more features are to include one or more of a maximum value, a minimum value, a baseline value, a linear acceleration integration, a Euler angle integration, an audio frequency spectrum or a post-collision vibration frequency spectrum, and

compare the one or more features to training data to classify the object.

11. The apparatus of claim 9 , wherein the logic is to trigger one or more of a competitive score update or an audiovisual effect based on whether the object is classified as the person or the second handheld device.

12. The apparatus of claim 8 , wherein the logic is to: detect an availability of one or more new data samples, and retrieve the motion data and the audio data from one or more buffers in response to the availability.

13. The apparatus of claim 8 , wherein the logic is to:

determine, based on the motion data, that a difference between a peak acceleration and an average acceleration has reached a first threshold, and

determine, based on the audio data, that a difference between a peak audio level and an average audio level has reached a second threshold.

14. A method comprising:

obtaining motion data;

obtaining audio data;

detecting a collision between a handheld device and an object based on the motion data and the audio data, and

classifying the object based on filtered collision data, wherein the object to be classified is one of a person or a second handheld device, and wherein the classification of the object includes: distinguishing between a first audible sound generation quality of a first strike event and a different second audible sound generation quality of a second strike event, wherein the first strike event is of the handheld device contacting the second handheld device and the second strike event is of the handheld device contacting the person, wherein the first strike event has the first audible sound generation quality based at least in part on a material of the second handheld device, and wherein the second strike event has the different second audible sound generation quality based at least in part on a material of the person.

15. The method of claim 14 , further including:

removing noise from the motion data and the audio data to obtain the filtered collision data.

16. The method of claim 15 , wherein classifying the object includes: identifying one or more features of the collision, wherein the one or more features include one or more of a maximum value, a minimum value, a baseline value, a linear acceleration integration, a Euler angle integration, an audio frequency spectrum or a post-collision vibration frequency spectrum; and

comparing the one or more features to training data.

17. The method of claim 15 , wherein the method further includes triggering one or more of a competitive score update or an audiovisual effect based on whether the object is classified as the person or the second handheld device.

18. The method of claim 14 , wherein obtaining the motion data and the audio data includes:

detecting an availability of one or more new data samples; and

retrieving the motion data and the audio data from one or more buffers in response to the availability.

19. The method of claim 14 , wherein detecting the collision includes:

determining, based on the motion data, that a difference between a peak acceleration and an average acceleration has reached a first threshold; and

determining, based on the audio data, that a difference between a peak audio level and an average audio level has reached a second threshold.

20. At least one computer readable storage medium comprising a set of instructions, which when executed by a handheld device, cause the handheld device to:

obtain motion data;

obtain audio data;

detect a collision between the handheld device and an object based on the motion data and the audio data, and

classify the object based on filtered collision data, wherein the object to be classified is one of a person or a second handheld device, and wherein the classification of the object includes: distinguishing between a first audible sound generation quality of a first strike event and a different second audible sound generation quality of a second strike event, wherein the first strike event is of the handheld device contacting the second handheld device and the second strike event is of the handheld device contacting the person, wherein the first strike event is to have the first audible sound generation quality based at least in part on a material of the second handheld device, and wherein the second strike event is to have the different second audible sound generation quality based at least in part on a material of the person.

21. The at least one computer readable storage medium of claim 20 , wherein the instructions, when executed, cause the handheld device to:

remove noise from the motion data and the audio data to obtain the filtered collision data.

22. The at least one computer readable storage medium of claim 21 , wherein the instructions, when executed, cause the handheld device to:

identify one or more features of the collision, wherein the one or more features are to include one or more of a maximum value, a minimum value, a baseline value, a linear acceleration integration, a Euler angle integration, an audio frequency spectrum or a post-collision vibration frequency spectrum; and

compare the one or more features to training data to classify the object.

23. The at least one computer readable storage medium of claim 21 , wherein the instructions, when executed, cause the handheld device to trigger one or more of a competitive score update or an audiovisual effect based on whether the object is classified as the person or the second handheld device.

24. The at least one computer readable storage medium of claim 20 , wherein the instructions, when executed, cause the handheld device to:

detect an availability of one or more new data samples; and

retrieve the motion data and the audio data from one or more buffers in response to the availability.

25. The at least one computer readable storage medium of claim 20 , wherein the instructions, when executed, cause the handheld device to:

determine, based on the motion data, that a difference between a peak acceleration and an average acceleration has reached a first threshold; and

determine, based on the audio data, that a difference between a peak audio level and an average audio level has reached a second threshold.