Vehicle entry detection

Patent metadata

Field	Value
Publication number	US-12148228-B2
Application number	US-201917274602-A
Country	US
Kind code	B2
Filing date	Oct 8, 2019
Priority date	Oct 8, 2018
Publication date	Nov 19, 2024
Grant date	Nov 19, 2024

Abstract

Certain aspects of the present disclosure are generally directed to apparatus and techniques for event state detection. One example method generally includes receiving a plurality of sensor signals at a computing device, determining, at the computing device, probabilities of sub-event states based on the plurality of sensor signals using an artificial neural network for each of a plurality of time intervals, and detecting, at the computing device, the event state based on the probabilities of the sub-event states via a state sequence model.

First claim

What is claimed is:

1. A method for event state detection, comprising:
receiving a plurality of sensor signals at a computing device;
determining, at the computing device, probabilities of sub-event states based on the plurality of sensor signals using an artificial neural network for each of a plurality of time intervals; and
detecting, at the computing device, an event state based on the probabilities of the sub-event states via a state sequence model.

2. The method of claim 1 , wherein detecting the event state comprises detecting a user entering a vehicle.

3. The method of claim 1 , wherein the plurality of sensor signals comprises an audio signal and at least one motion sensor signal.

4. The method of claim 3 , wherein the at least one motion sensor signal comprises at least one of an accelerometer sensor signal, a gyroscope sensor signal, or a magnetometer sensor signal.

5. The method of claim 1 , wherein the plurality of sensor signals comprises motion sensor signals.

6. The method of claim 1 , wherein the plurality of sensor signals comprises an audio signal.

7. The method of claim 1 , wherein determining the probabilities of the sub-event states comprises:
determining state probabilities corresponding to each of the plurality of sensor signals via a convolutional neural network (CNN); and
fusing the state probabilities via a deep neural network (DNN).

8. The method of claim 7 , wherein the CNN comprises three neural network layers.

9. The method of claim 7 , wherein the DNN comprises a single neural network layer.

10. The method of claim 1 , wherein the state sequence model comprises a hidden Markov model or a hidden semi-Markov model.

11. The method of claim 1 , wherein the detection of the event state comprises detecting a sequence of the sub-event states based on the determination of the probabilities.

12. The method of claim 11 , wherein the event state is detected based on the sequence and a duration of each of the sub-event states.

13. The method of claim 11 , wherein the sequence of the sub-event states comprises a sequence of 15 sub-event states.

14. The method of claim 1 , further comprising:
determining a level of confidence that the event state has occurred via the state sequence model, wherein the event state is detected when the level of confidence exceeds a threshold.

15. The method of claim 1 , further comprising extracting a feature from each of the plurality of sensor signals, wherein the probabilities of the sub-event states are determined based on the extracted features.

16. The method of claim 1 , wherein the computing device comprises a wearable or mobile device and wherein the plurality of sensor signals is received from sensors of the wearable or mobile device.

17. The method of claim 1 , further comprising:
time synchronizing the plurality of sensor signals, wherein the determination of the probabilities of the sub-event states is based on the time-synchronized sensor signals.

18. The method of claim 17 , further comprising:
resampling each of the time-synchronized sensor signals such that the time-synchronized sensor signals have the same sampling rate, wherein the determination of the probabilities of sub-event states is based on the resampled time-synchronized sensor signals.

19. The method of claim 1 , further comprising performing an action based on the detection of the event state.

20. The method of claim 19 , wherein performing the action comprises at least one of:
activating a navigation system;
displaying information for a user; or
transmitting an indication to a vehicle to perform one or more actions.

21. The method of claim 1 , wherein the plurality of sensor signals comprises at least two different types of sensor signals.

22. An apparatus for event state detection, comprising:
a plurality of sensors; and
a processing system configured to:
receive a plurality of sensor signals from the plurality of sensors;
determine probabilities of sub-event states based on the plurality of sensor signals using an artificial neural network for each of a plurality of time intervals; and
detect an event state based on the probabilities of the sub-event states via a state sequence model.

23. The apparatus of claim 22 , wherein the processing system is configured to detect the event state by detecting a user entering a vehicle.

24. The apparatus of claim 22 , wherein the processing system is further configured to determine a level of confidence that the event state has occurred via the state sequence model, wherein the event state is detected when the level of confidence exceeds a threshold.

25. The apparatus of claim 22 , wherein the processing system is further configured to extract a feature from each of the plurality of sensor signals, wherein the probabilities of the sub-event states are determined based on the extracted features.

26. The apparatus of claim 22 , wherein the processing system is further configured to time synchronize the plurality of sensor signals, wherein the determination of the probabilities of the sub-event states is based on the time-synchronized sensor signals.

27. The apparatus of claim 26 , wherein the processing system is further configured to resample each of the time-synchronized sensor signals such that the time-synchronized sensor signals have the same sampling rate, wherein the determination of the probabilities of sub-event states is based on the resampled time-synchronized sensor signals.

28. The apparatus of claim 22 , wherein the processing system is further configured to perform an action based on the detection of the event state.

29. An apparatus for event state detection, comprising:
means for receiving a plurality of sensor signals;
means for determining probabilities of sub-event states based on the plurality of sensor signals using an artificial neural network for each of a plurality of time intervals; and
means for detecting an event state based on the probabilities of the sub-event states via a state sequence model.

30. A non-transitory computer-readable medium having instructions stored thereon to cause a computing device to:
receive a plurality of sensor signals;
determine probabilities of sub-event states based on the plurality of sensor signals using an artificial neural network for each of a plurality of time intervals; and
detect an event state based on the probabilities of the sub-event states via a state sequence model.

Assignees

• Qualcomm Inc

Inventors

• Lee Mingu
• Chang Wonil
• Kim Yeonseok
• Hwang Kyu Woong
• Huang Yin
• Wang Ruowei
• Zhao Haijun
• Cho Janghoon

Classifications

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06V10/764

  using classification, e.g. of video objects · CPC title

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06N3/02

  Neural networks · CPC title