Face recognition systems with external stimulus
An apparatus includes a camera and a processor. The camera may be configured to capture video data. The processor may be configured to (A) process the video data, (B) generate control signals used to initiate a stimulus and (C) execute computer readable instructions. The computer readable instructions may be executed by the processor to perform video analysis on video frames of the captured video data to (a) detect a person, (b) detect context information associated with the detected person and (c) determine facial recognition results of the detected person. If the facial recognition results cannot detect a face of the detected person, the processor selects the stimulus from a plurality of stimuli. The stimulus may be selected in response to the context information to increase a probability of detecting a frontal view of the face of the detected person.
1. An apparatus comprising:
a camera sensor configured to capture video data; and
a processor configured to (A) process said video data, (B) generate control signals used to initiate a stimulus and (C) execute computer readable instructions, wherein (i) said computer readable instructions are executed by said processor to perform video analysis on video frames of said captured video data to (a) detect a person, (b) detect context information associated with said detected person and (c) determine facial recognition results of said detected person, (ii) if said facial recognition results cannot detect a face of said detected person, said processor selects said stimulus from a plurality of stimuli and (iii) said stimulus is selected in response to said context information to increase a probability of detecting a frontal view of said face of said detected person.
2. The apparatus according to claim 1 , wherein said probability of detecting said frontal view of said face of said detected person is increased by encouraging said detected person to look at said camera sensor.
3. The apparatus according to claim 1 , wherein said facial recognition results are determined in real-time.
4. The apparatus according to claim 3 , wherein said real-time for said facial recognition results is within 500 ms.
5. The apparatus according to claim 1 , wherein said processor is configured to share video processing tasks with a cloud server based on a processing capability of said processor.
6. The apparatus according to claim 5 , wherein said processor and said cloud server are configured to implement a video processing pipeline with different video processing tasks performed by said processor and said cloud server.
7. An apparatus comprising:
a camera sensor configured to capture video data; and
a processor configured to (A) process said video data, (B) generate control signals used to initiate a stimulus and (C) execute computer readable instructions, wherein (i) said computer readable instructions are executed by said processor to (a) stream said video data to an external server, (b) receive (A) facial recognition results and (B) context information associated with a detected person from said external server corresponding to said video data and (c) select said stimulus from a plurality of stimuli, (ii) if said facial recognition results cannot detect a face of said detected person, said stimulus is selected from said plurality of stimuli and (iii) said stimulus is selected in response to said context information to increase a probability of detecting a frontal view of said face of said detected person.
8. The apparatus according to claim 7 , wherein said apparatus is configured to (i) detect motion and (ii) if said motion is detected, generate said control signals to initiate said stimulus.
9. The apparatus according to claim 7 , wherein person detection operations are performed by said processor to detect said detected person as a filter before facial recognition is implemented.
10. The apparatus according to claim 7 , wherein (i) said video data captured during a pre-determined time after said stimulus is initiated is used as selected video frames and (ii) said selected video frames are streamed to said external server.
11. The apparatus according to claim 10 , wherein (i) said selected video frames have a higher probability of having said frontal view of said face of said detected person and (ii) streaming only said selected frames increases a speed of generating said facial recognition results.
12. The apparatus according to claim 7 , wherein said facial recognition results are presented to a user device for evaluation by a user.
13. The apparatus according to claim 12 , wherein (i) said evaluation by said user is presented to said external server as feedback and (ii) said feedback is used to improve learning for video processing tasks.
14. The apparatus according to claim 7 , wherein said external server comprises a distributed computing network configured to scale resources based on demand.
15. The apparatus according to claim 7 , wherein said plurality of stimuli comprise at least one of (a) non-interactive stimulus, (b) conversation stimulus, (c) directive stimulus and (d) alarm stimulus.
16. The apparatus according to claim 15 , wherein said non-interactive stimulus comprises an unexpected sound.
17. The apparatus according to claim 15 , wherein said conversation stimulus comprises an interactive conversation using a two-way radio functionality of said apparatus between at least one of (i) said detected person and a homeowner, (ii) said detected person and an artificial intelligence and (iii) said detected person and a combination of said homeowner and said artificial intelligence.
18. The apparatus according to claim 15 , wherein said directive stimulus comprises instructing said detected person to perform a task that comprises facing said camera sensor.
19. The apparatus according to claim 15 , wherein said alarm stimulus comprises a distress signal.
20. An apparatus comprising:
a camera sensor configured to capture video data; and
a processor configured to (A) process said video data, (B) generate control signals used to initiate a stimulus and (C) execute computer readable instructions, wherein (i) said computer readable instructions are executed by said processor to perform video analysis on video frames of said captured data to (a) detect a person, (b) detect context information associated with said detected person and (c) determine whether said detected person has performed a desired response, (ii) if said desired response of said detected person has not been detected, said processor selects said stimulus from a plurality of stimuli and (iii) said stimulus that is selected has a highest probability out of said plurality of stimuli of causing said desired response from said detected person.