Dynamic image recognition and training using data center resources and data
A system, method, and computer-readable medium are disclosed for creating image recognition models, which can be operated on smartphone or similar device. The smartphone captures images of hardware in a data center. The captured images are processed to produce a full set of annotated images. The full set is minimized to a simplified set and trained to create a mobile image recognition model implemented by the smartphone or similar device.
1. A computer-implementable method for creating image recognition models, the method comprising:
receiving captured images of hardware in a data center;
processing the captured images to produce a full set of annotated captured images;
minimizing the full set of annotated captured images with a template, wherein a minimum set of classifications based on pruning classifications in a master set of annotations is used to create a simplified set of images, wherein the pruning classifications is performed by identifying annotation patterns and applying multiple heuristics; and
training the simplified set of images into a mobile image recognition model, to understand patterns on the original annotated captured images to predict location of the hardware in the data center.
2. The method of claim 1 , wherein the captured images are image frames or video captured by a smartphone implementing Bluetooth Low Energy (BLE).
3. The method of claim 1 , wherein the captured images include location in the data center.
4. The method of claim 1 , wherein processing is done in sequence with a node or server in the data center and/or a smartphone.
5. The method of claim 1 , wherein the processing includes receiving periodically a new or updated training master image recognition model from a master node or server in the data center.
6. The method of claim 1 , wherein the processing includes identifying available nodes or servers, and processors in the available nodes and servers, and queuing the nodes or servers and processors.
7. The method of claim 1 , wherein the training includes receiving periodically new or updated training data from a master node or server in the data center.
8. A system comprising:
a processor;
a data bus coupled to the processor; and
a non-transitory, computer-readable storage medium embodying computer program code, the non-transitory, computer-readable storage medium being coupled to the data bus, the computer program code interacting with a plurality of computer operations for creating image recognition models and comprising instructions executable by the processor and configured for:
receiving captured images of hardware in a data center;
processing the captured images to produce a full set of annotated captured images;
minimizing the full set of annotated captured images with a template, wherein a minimum set of classifications based on pruning classifications in a master set of annotations is used to create a simplified set of images, wherein the pruning classifications is performed by identifying annotation patterns and applying multiple heuristics; and
training the simplified set of images into a mobile image recognition model, to understand patterns on the original annotated captured images to predict location of the hardware in the data center.
9. The system of claim 8 , wherein the captured images are image frames or video captured by a smartphone implementing Bluetooth Low Energy (BLE).
10. The system of claim 8 , wherein the captured images include location in the data center.
11. The system of claim 8 , wherein processing is done in sequence with a node or server in the data center and/or a smartphone.
12. The system of claim 8 , wherein the processing includes receiving periodically a new or updated training master image recognition model from a master node or server in the data center.
13. The system of claim 8 , wherein the processing includes identifying available nodes or servers, and processors in the available nodes and servers, and queuing the nodes or servers and processors.
14. The system of claim 8 , wherein the training includes receiving periodically new or updated training data from a master node or server in the data center.
15. A non-transitory, computer-readable storage medium embodying computer program code, the computer program code comprising computer executable instructions configured for:
receiving captured images of hardware in a data center;
processing the captured images to produce a full set of annotated captured images;
minimizing the full set of annotated captured images with a template, wherein a minimum set of classifications based on pruning classifications in a master set of annotations is used to create a simplified set of images, wherein the pruning classifications is performed by identifying annotation patterns and applying multiple heuristics; and
training the simplified set of images into a mobile image recognition model, to understand patterns on the original annotated captured images to predict location of the hardware in the data center.
16. The non-transitory, computer-readable storage medium of claim 15 , wherein the captured images include location in the data center.
17. The non-transitory, computer-readable storage medium of claim 15 , wherein the processing is done in sequence with a node or server in the data center and/or a smartphone implementing Bluetooth Low Energy (BLE).
18. The non-transitory, computer-readable storage medium of claim 15 , wherein the processing includes receiving periodically a new or updated training master image recognition model from a master node or server in the data center.
19. The non-transitory, computer-readable storage medium of claim 15 , wherein the processing includes identifying available nodes or servers, and processors in the available nodes and servers, and queuing the nodes or servers and processors.
20. The non-transitory, computer-readable storage medium of claim 15 , wherein the training includes receiving periodically new or updated training data from a master node or server in the data center.