# US 2021/0264163 A1 · Person tracking and privacy and acceleration of data using autonomous machines

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2021/0264163 A1
- Document ID: US20210264163A1
- Status: Published application
- Kind: A1
- Application number: 17/173,732
- Filed: 2021-02-11
- Published: 2021-08-26
- Earliest claimed: 2017-04-17
- Assignees: Intel Corporation
- Applicants: Intel Corporation
- Inventors: Mayuresh M. Varerkar; Barnan Das; Narayan Biswal; Stanley J. Baran; Gokcen Cilingir; Nilesh V. Shah; Archie Sharma; Sherine Abdelhak; Sachin Godse; Farshad Akhbari; Narayan Srinivasa; Altug Koker; Nadathur Rajagopalan Satish; Dukhwan Kim; Feng Chen; Abhishek R. Appu; Joydeep Ray; Ping T. Tang; Michael S. Strickland; Xiaoming Chen; Anbang Yao; Tatiana Shpeisman; Vasanth Ranganathan; Sanjeev Jahagirdar
- Web page: https://patentgrep.com/patent/US20210264163A1

## CPC

- G06V 20/52: Surveillance or monitoring of activities, e.g. for recognising suspicious objects (recognising microscopic objects G06V20/69)
- G06N 3/044: Recurrent networks, e.g. Hopfield networks
- G06N 3/0442: characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU]
- G06N 3/045: Combinations of networks
- G06N 3/0464: Convolutional networks [CNN, ConvNet]
- G06N 3/048: Activation functions
- G06N 3/063: using electronic means
- G06N 3/084: Backpropagation, e.g. using gradient descent
- G06N 3/0895: Weakly supervised learning, e.g. semi-supervised or self-supervised learning
- G06N 3/09: Supervised learning
- G06N 3/098: Distributed learning, e.g. federated learning
- G06T 1/20: Processor architectures; Processor configuration, e.g. pipelining
- G06V 20/40: in video content (extracting overlay text G06V20/62; video retrieval G06F16/70; processing of video elementary streams in video servers H04N21/234; processing of video elementary streams in video clients H04N21/44)
- G06V 40/10: Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V 20/44: {Event detection}
- G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL

## Abstract

A mechanism is described for facilitating person tracking and data security in machine learning at autonomous machines. A method of embodiments, as described herein, includes detecting, by a camera associated with one or more trackers, a person within a physical vicinity, where detecting includes capturing one or more images the person. The method may further include tracking, by the one or more trackers, the person based on the one or more images of the person, where tracking includes collect tracking data relating to the person. The method may further include selecting a tracker of the one or more trackers as a preferred tracker based on the tracking data.

## Claims

### Claim at position 1

1-20. (canceled)

### Claim 21

An apparatus comprising:
one or more processors including a graphics processing unit (GPU);
a memory to store data, the data including data for a neural network; and
one or more field programmable gate array (FPGA) units;
wherein the apparatus is to:
receive application code for compilation, the application including a machine learning operation;
identify one or more mathematics operations to be mapped to the one or more FPGA units;
detecting one or more instances of the identified mathematics operations in the received application code; and
generating compiled code for the application, wherein generating the compiled code includes replacing the detected one or more instances of the identified mathematics operations with send messages to access blocks of the one or more FPGA units.

### Claim 22 (depends on 21)

The apparatus of claim 21, wherein identifying the one or more mathematics operations to be mapped to the one or more FPGA units including determining the one or more mathematics operations from the application.

### Claim 23 (depends on 21)

The apparatus of claim 21, wherein the identified mathematics operations include operations for deep learning to be accelerated by the one or more FPGA units.

### Claim 24 (depends on 23)

The apparatus of claim 23, wherein the identified mathematical operations include one or more commonly used math functions to allow for low power and high performance.

### Claim 25 (depends on 23)

The apparatus of claim 23, wherein the identified mathematical operations include an activation function for deep learning.

### Claim 26 (depends on 21)

The apparatus of claim 21, wherein the blocks of the one or more FPGA units are distributed among a plurality of elements of the GPU.

### Claim 27 (depends on 26)

The apparatus of claim 26, wherein the blocks of the one or more FPGA units are distributed among a plurality of streaming multiprocessors (SMs) of the GPU.

### Claim 28 (depends on 26)

The apparatus of claim 26, wherein the blocks of the one or more FPGA units are distributed among a plurality of slices or cores of the GPU.

### Claim 29

At least one non-transitory machine-readable medium comprising instructions that when executed by a computing device, cause the computing device to perform operations comprising:
receiving an application for compilation, the application including a machine learning operation;
identifying a set of one or more mathematics operations to be mapped to one or more field programmable gate array (FPGA) units in a system, the system further including a graphics processing unit (GPU);
detecting one or more instances of the set of mathematics operations in the application to be mapped to the FPGA units; and
generating code for operation of the application, wherein generating the code includes replacing the one or more instances of the set of mathematics operations with send messages to access blocks of the one or more FPGA units.

### Claim 30 (depends on 29)

The machine-readable medium of claim 29, wherein identifying the one or more mathematics operations to be mapped to the one or more FPGA units including determining the one or more mathematics operations from the application.

### Claim 31 (depends on 29)

The machine-readable medium of claim 29, wherein the identified mathematics operations include operations for deep learning to be accelerated by the one or more FPGA units.

### Claim 32 (depends on 31)

The machine-readable medium of claim 31, wherein the identified mathematical operations include one or more commonly used math functions to allow for low power and high performance.

### Claim 33 (depends on 31)

The machine-readable medium of claim 31, wherein the identified mathematical operations include an activation function for deep learning.

### Claim 34

A computing system comprising:
one or more processors including a graphics processing unit (GPU);
a memory to store data, the data including data for a neural network; and
one or more field programmable gate array (FPGA) units, the GPU and the one or more FPGA units forming a hybrid system for deep learning;
wherein the computing system is to:
receive application code for compilation, the application including a deep learning operation;
identify one or more mathematics operations to be mapped to the one or more FPGA units;
detecting one or more instances of the identified mathematics operations in the received application code; and
generating compiled code for the application, wherein generating the compiled code includes replacing the detected one or more instances of the identified mathematics operations with send messages to access blocks of the one or more FPGA units.

### Claim 35 (depends on 34)

The computing system of claim 34, wherein:
the one or more FPGA units are to accelerate functions used in deep learning, and
the GPU is to be applied at least in part for one or more general purpose tasks and algorithms.

### Claim 36 (depends on 34)

The computing system of claim 34, wherein identifying the one or more mathematics operations to be mapped to the one or more FPGA units includes obtaining hints regarding mathematics operations from the application.

### Claim 37 (depends on 34)

The computing system of claim 34, wherein the identified mathematical operations include one or more commonly used math functions to allow for low power and high performance.

### Claim 38 (depends on 37)

The computing system of claim 37, wherein the identified mathematical operations include an activation function for deep learning.

### Claim 39 (depends on 34)

The computing system of claim 34, wherein the blocks of the one or more FPGA units are distributed among a plurality of elements of the GPU.

### Claim 40 (depends on 39)

The computing system of claim 39, wherein the blocks of the one or more FPGA units are distributed among:
a plurality of streaming multiprocessors (SMs) of the GPU; or
a plurality of slices or cores of the GPU.
