Vehicular system for testing performance of headlamp detection systems

1. A method for testing a vehicular driving assist system, the method comprising:

providing a neural network;

training the neural network using a database of images, wherein each image of the database comprises an image of a headlight of a vehicle or an image of a taillight of a vehicle;

providing to the trained neural network an input image that does not include a headlight or a taillight;

generating, by the neural network using the input image, an output image, wherein the output image comprises an image of a headlight generated by the neural network or an image of a taillight generated by the neural network; and

testing the vehicular driving assist system using the generated output image as an input to the vehicular driving assist system.

2. The method of claim 1 , wherein the neural network comprises a generative adversarial network.

3. The method of claim 2 , wherein training the neural network comprises iterating the training until a discriminator cannot distinguish between the input image and the generated output image.

4. The method of claim 1 , wherein the vehicular driving assist system comprises an automatic headlight control system.

5. The method of claim 1 , wherein the database comprises a plurality of video recordings.

6. The method of claim 5 , wherein the plurality of video recordings comprises video recordings of a headlight or a taillight at at least one selected from the group consisting of (i) different speeds relative to a camera capturing the video recordings, (ii) different distances relative to the camera capturing the video recordings and (iii) different orientations relative to the camera capturing the video recordings.

7. The method of claim 1 , wherein the vehicular driving assist system classifies the generated output image as including a headlight or a taillight and determines a position of the headlight or the taillight within the output image.

8. The method of claim 1 , wherein the database comprises GPS data corresponding to the images.

9. The method of claim 8 , wherein the GPS data emulates an effect of speed and distance experienced by a camera capturing the images.

10. The method of claim 1 , wherein the neural network generates the output image by modifying the input image with the image of the headlight or the taillight.

11. A method for testing a vehicular driving assist system, the method comprising:

providing a generative adversarial network (GAN);

training the GAN using a database of video recordings, wherein each video recording of the database comprises a plurality of images that include a headlight of a vehicle or a taillight of a vehicle;

providing to the trained GAN an input image that does not include a headlight or a taillight;

generating, by the GAN using the input image, an output image, wherein the output image comprises an image of a headlight generated by the GAN or an image of a taillight generated by the GAN; and

testing the vehicular driving assist system using the generated output image as an input to the vehicular driving assist system.

12. The method of claim 11 , wherein training the GAN comprises iterating the training until a discriminator cannot distinguish between the input image and the generated output image.

13. The method of claim 11 , wherein the vehicular driving assist system comprises an automatic headlight control system.

14. The method of claim 11 , wherein the database of video recordings comprises video recordings of a headlight or a taillight at at least one selected from the group consisting of (i) different speeds relative to a camera capturing the video recordings, (ii) different distances relative to the camera capturing the video recordings and (iii) different orientations relative to the camera capturing the video recordings.

15. The method of claim 11 , wherein the vehicular driving assist system classifies the output image as including a headlight or a taillight and determines a position of the headlight or the taillight within the output image.

16. The method of claim 11 , wherein the database comprises GPS data corresponding to the video recordings.

17. A method for testing an automatic headlight control system, the method comprising:

providing a neural network;

training the neural network using a database of images, wherein each image of the database comprises an image of a headlight of a vehicle or an image of a taillight of a vehicle;

providing to the trained neural network an input image that does not include a headlight or a taillight;

generating, by the neural network using the input image, an output image, wherein the output image comprises an image of a headlight generated by the neural network or an image of a taillight generated by the neural network;

testing the vehicular driving assist system using the generated output image as an input to the automatic headlight control system; and

wherein the automatic headlight control system classifies the generated output image as including a headlight or a taillight and determines a position of the headlight or the taillight within the generated output image.

18. The method of claim 17 , wherein the neural network comprises a generative adversarial network.

19. The method of claim 18 , wherein training the neural network comprises iterating the training until a discriminator cannot distinguish between the input image and the generated output image.

20. The method of claim 17 , wherein the database comprises a plurality of video recordings.