Ubiquitous visual computing witness
In one embodiment, an apparatus comprises a memory and a processor. The memory is to store sensor data captured by one or more sensors associated with a first device. Further, the processor comprises circuitry to: access the sensor data captured by the one or more sensors associated with the first device; determine that an incident occurred within a vicinity of the first device; identify a first collection of sensor data associated with the incident, wherein the first collection of sensor data is identified from the sensor data captured by the one or more sensors; preserve, on the memory, the first collection of sensor data associated with the incident; and notify one or more second devices of the incident, wherein the one or more second devices are located within the vicinity of the first device.
1. A computing device for reconstructing anomalous events from sensor data captured by edge devices, comprising:
communication circuitry to communicate over a network; and
processing circuitry to:
determine that an anomalous event occurred;
send, via the communication circuitry, a request for sensor data captured within a proximity of the anomalous event;
receive, via the communication circuitry, multimodal sensor data from a plurality of edge devices within the proximity of the anomalous event, wherein the multimodal sensor data is captured by a plurality of types of sensors from a plurality of perspectives; and
generate, based on the multimodal sensor data, a multidimensional representation of the anomalous event, wherein the multidimensional representation of the anomalous event is to represent the anomalous event from the plurality of perspectives captured by the plurality of types of sensors.
2. The computing device of claim 1 , wherein the processing circuitry to determine that the anomalous event occurred is further to:
receive, via the communication circuitry, an indication that the anomalous event occurred from a first edge device within the proximity of the anomalous event.
3. The computing device of claim 1 , wherein the processing circuitry to determine that the anomalous event occurred is further to:
obtain a first set of sensor data captured within the proximity of the anomalous event; and
detect the anomalous event based on an analysis of the first set of sensor data.
4. The computing device of claim 3 , wherein the processing circuitry to obtain the first set of sensor data captured within the proximity of the anomalous event is further to:
receive, via the communication circuitry, the first set of sensor data from a first edge device, wherein the first set of sensor data is captured within the proximity of the anomalous event by the first edge device.
5. The computing device of claim 3 , wherein:
the computing device further comprises one or more sensors to capture the first set of sensor data within the proximity of the anomalous event; and
the processing circuitry to obtain the first set of sensor data captured within the proximity of the anomalous event is further to:
obtain the first set of sensor data from the one or more sensors.
6. The computing device of claim 3 , wherein the processing circuitry to send, via the communication circuitry, the request for sensor data captured within the proximity of the anomalous event is further to:
identify, based on the first set of sensor data, the plurality of edge devices within the proximity of the anomalous event; and
send, via the communication circuitry, the request for sensor data to the plurality of edge devices within the proximity of the anomalous event.
7. The computing device of claim 1 , wherein the processing circuitry to send, via the communication circuitry, the request for sensor data captured within the proximity of the anomalous event is further to:
broadcast the request for sensor data over an information-centric network, wherein the request for sensor data is to indicate a plurality of attributes associated with the anomalous event, wherein the plurality of attributes comprise a time window and a geographical region of interest corresponding to the anomalous event.
8. The computing device of claim 1 , wherein the processing circuitry to generate, based on the multimodal sensor data, the multidimensional representation of the anomalous event is further to:
perform sensor fusion on the multimodal sensor data to generate the multidimensional representation of the anomalous event from the plurality of perspectives captured by the plurality of types of sensors.
9. The computing device of claim 1 , wherein the processing circuitry is further to:
receive, via the communication circuitry, a request for information associated with the anomalous event; and
send, via the communication circuitry and based on the request for information associated with the anomalous event, at least one of:
the multidimensional representation of the anomalous event; or
the plurality of sets of sensor data.
10. The computing device of claim 1 , wherein:
the anomalous event is associated with an automobile accident; and
the computing device is associated with:
a vehicle within the proximity of the automobile accident; or
roadside infrastructure within the proximity of the automobile accident.
11. The computing device of claim 1 , wherein:
the anomalous event is associated with criminal activity; and
the computing device is associated with a surveillance system within the proximity of the criminal activity.
12. The computing device of claim 1 , wherein the proximity of the anomalous event comprises a geographical proximity and a temporal proximity.
13. At least one non-transitory machine accessible storage medium having instructions stored thereon, wherein the instructions, when executed on processing circuitry, cause the processing circuitry to:
determine that an anomalous event occurred;
send, via communication circuitry for communicating over a network, a request for sensor data captured within a proximity of the anomalous event;
receive, via the communication circuitry, multimodal sensor data from a plurality of edge devices within the proximity of the anomalous event, wherein the multimodal sensor data is captured by a plurality of types of sensors from a plurality of perspectives; and
generate, based on the multimodal sensor data, a multidimensional representation of the anomalous event, wherein the multidimensional representation of the anomalous event is to represent the anomalous event from the plurality of perspectives captured by the plurality of types of sensors.
14. The storage medium of claim 13 , wherein the instructions that cause the processing circuitry to determine that the anomalous event occurred further cause the processing circuitry to:
obtain a first set of sensor data captured within the proximity of the anomalous event; and
detect the anomalous event based on an analysis of the first set of sensor data.
15. The storage medium of claim 14 , wherein the instructions that cause the processing circuitry to obtain the first set of sensor data captured within the proximity of the anomalous event further cause the processing circuitry to:
obtain the first set of sensor data from one or more sensors; or
receive, via the communication circuitry, the first set of sensor data from a first edge device, wherein the first set of sensor data is captured within the proximity of the anomalous event by the first edge device.
16. The storage medium of claim 14 , wherein the instructions that cause the processing circuitry to send, via the communication circuitry, the request for sensor data captured within the proximity of the anomalous event further cause the processing circuitry to:
identify, based on the first set of sensor data, the plurality of edge devices within the proximity of the anomalous event; and
send, via the communication circuitry, the request for sensor data to the plurality of edge devices within the proximity of the anomalous event.
17. The storage medium of claim 13 , wherein the instructions that cause the processing circuitry to send, via the communication circuitry, the request for sensor data captured within the proximity of the anomalous event further cause the processing circuitry to:
broadcast the request for sensor data over an information-centric network, wherein the request for sensor data is to indicate a plurality of attributes associated with the anomalous event, wherein the plurality of attributes comprise a time window and a geographical region of interest corresponding to the anomalous event.
18. The storage medium of claim 13 , wherein the instructions that cause the processing circuitry to generate, based on the multimodal sensor data, the multidimensional representation of the anomalous event further cause the processing circuitry to:
perform sensor fusion on the multimodal sensor data to generate the multidimensional representation of the anomalous event from the plurality of perspectives captured by the plurality of types of sensors.
19. The storage medium of claim 13 , wherein the instructions further cause the processing circuitry to:
receive, via the communication circuitry, a request for information associated with the anomalous event; and
send, via the communication circuitry and based on the request for information associated with the anomalous event, at least one of:
the multidimensional representation of the anomalous event; or
the plurality of sets of sensor data.
20. The storage medium of claim 13 , wherein:
the anomalous event is associated with an automobile accident; and
the processing circuitry is associated with:
a vehicle within the proximity of the automobile accident; or
roadside infrastructure within the proximity of the automobile accident.
21. The storage medium of claim 13 , wherein:
the anomalous event is associated with criminal activity; and
the processing circuitry is associated with a surveillance system within the proximity of the criminal activity.
22. A method of reconstructing anomalous events from sensor data captured by edge devices, comprising:
determining that an anomalous event occurred;
sending, via communication circuitry for communicating over a network, a request for sensor data captured within a proximity of the anomalous event;
receiving, via the communication circuitry, multimodal sensor data from a plurality of edge devices within the proximity of the anomalous event, wherein the multimodal sensor data is captured by a plurality of types of sensors from a plurality of perspectives; and
generating, based on the multimodal sensor data, a multidimensional representation of the anomalous event, wherein the multidimensional representation of the anomalous event is to represent the anomalous event from the plurality of perspectives captured by the plurality of types of sensors.
23. The method of claim 22 , wherein:
determining that the anomalous event occurred comprises:
obtaining a first set of sensor data captured within the proximity of the anomalous event; and
detecting the anomalous event based on an analysis of the first set of sensor data; and
sending, via the communication circuitry, the request for sensor data captured within the proximity of the anomalous event comprises:
identifying, based on the first set of sensor data, the plurality of edge devices within the proximity of the anomalous event; and
sending, via the communication circuitry, the request for sensor data to the plurality of edge devices within the proximity of the anomalous event.
24. The method of claim 22 , wherein sending, via the communication circuitry, the request for sensor data captured within the proximity of the anomalous event comprises:
broadcasting the request for sensor data over an information-centric network, wherein the request for sensor data is to indicate a plurality of attributes associated with the anomalous event, wherein the plurality of attributes comprise a time window and a geographical region of interest corresponding to the anomalous event.
25. The method of claim 22 , wherein generating, based on the multimodal sensor data, the multidimensional representation of the anomalous event comprises:
performing sensor fusion on the multimodal sensor data to generate the multidimensional representation of the anomalous event from the plurality of perspectives captured by the plurality of types of sensors.
26. The computing device of claim 1 , wherein the plurality of perspectives comprises:
perspectives from different camera views; or
perspectives from different sensor types.