REAL-TIME OBJECT DETECTION FROM DECOMPRESSED IMAGES
A system and method for deploying machine learning models with consistent fixed-point arithmetic processing. A computing device in a vehicle receives a trained machine learning model that was trained using training images processed with fixed-point arithmetic operations. The computing device receives images from a camera and processes the received images using fixed-point arithmetic operations that are consistent with those used during training. The machine learning model is executed using the processed images to detect objects such as vehicles, pedestrians, persons on bikes, stop lights, or stop signs. The computing device may comprise specialized hardware including a graphics processing unit (GPU), digital signal processing hardware decoder, or single-purpose hardware decoder such as an application-specific integrated circuit (ASIC) or field-programmable gate array (FPGA). The system may generate alerts or determine vehicle operation changes based on detected objects, and may operate in surveillance mode when the vehicle is parked.
1 . A method comprising:
receiving, at a computing device in a vehicle, a trained machine learning model, wherein the machine learning model was trained using training images processed with fixed-point arithmetic operations consistent with fixed-point arithmetic operations of the computing device;
receiving, by the computing device, images from a camera;
processing the received images using the fixed-point arithmetic operations of the computing device to generate processed images; and
executing the trained machine learning model using the processed images.
2 . The method of claim 1 , wherein the fixed-point arithmetic operations of the computing device comprise a resizing operation.
3 . The method of claim 1 , wherein the computing device comprises a graphics processing unit (GPU).
4 . The method of claim 1 , wherein the computing device comprises a digital signal processing hardware decoder.
5 . The method of claim 1 , wherein the computing device comprises a single-purpose hardware decoder.
6 . The method of claim 5 , wherein the single-purpose hardware decoder comprises an application-specific integrated circuit (ASIC) or a field-programmable gate array (FPGA).
7 . The method of claim 1 , wherein executing the machine learning model comprises detecting an object within the processed images, and wherein a detected object comprises at least one of: a vehicle, a pedestrian, a person on a bike, a stop light, or a stop sign.
8 . The method of claim 7 , further comprising generating an alert based on the detected object.
9 . The method of claim 7 , further comprising determining a change in vehicle operation based on the detected object.
10 . The method of claim 1 , wherein the computing device operates in a surveillance mode when the vehicle is parked.
11 . A system comprising:
a computing device in a vehicle, the computing device comprising a processor configured to:
receive a trained machine learning model, wherein the machine learning model was trained using training images processed with fixed-point arithmetic operations consistent with fixed-point arithmetic operations of the computing device;
receive images from a camera;
process the received images using the fixed-point arithmetic operations of the computing device to generate processed images; and
execute the trained machine learning model using the processed images.
12 . The system of claim 11 , wherein the fixed-point arithmetic operations of the computing device comprise a resizing operation.
13 . The system of claim 11 , wherein the computing device comprises a graphics processing unit (GPU).
14 . The system of claim 11 , wherein the computing device comprises a digital signal processing hardware decoder.
15 . The system of claim 11 , wherein the computing device comprises a single-purpose hardware decoder.
16 . The system of claim 15 , wherein the single-purpose hardware decoder comprises an application-specific integrated circuit (ASIC).
17 . The system of claim 11 , wherein the processor is configured to execute the machine learning model to detect an object within the processed images.
18 . The system of claim 17 , wherein a detected object comprises at least one of: a vehicle, a pedestrian, a person on a bike, a stop light, or a stop sign.
19 . The system of claim 17 , wherein the processor is further configured to generate an alert based on the detected object.
20 . The system of claim 11 , wherein the computing device is configured to operate in a surveillance mode when the vehicle is parked.