Uriel Rosa - Revision history

imported>Student221: /* Methods */

2020-11-24T01:18:33Z

Methods

← Older revision		Revision as of 01:18, 24 November 2020
Line 115:		Line 115:
	'''Resnet50 network'''		'''Resnet50 network'''

	A total of four distinct classes were formed through the simulated pipelines for each of the effects to be investigated: monochrome, defocused monochrome, RGB and defocused RGB. The modifications were introduced in the default pipeline to obtain these four simulated pipelines. The Resnet50 network upper layers were retrained with the original and the new images by following the MLTransferLearning guidelines [13]. A total of 50 pipeline processed images per class were enhanced to produce another 50 images for the CNN classifications by randomly reflecting these images around the vertical x of the frame and scaling from a factor between 1 and 2 [figure 8]. The assessment of the capability of each pipeline for correctly classifying the images was used as comparison criteria. The classification performance by training the original images was adopted as a baseline. The initial learning rate of the network was adjusted to 0.0001.		A total of four distinct classes were formed through the simulated pipelines for each of the effects to be investigated: monochrome, defocused monochrome, RGB and defocused RGB. The modifications were introduced in the default pipeline to obtain these four simulated pipelines. The Resnet50 network upper layers were retrained with the original and the new images by following the MLTransferLearning guidelines [14]. A total of 50 pipeline processed images per class were enhanced to produce another 50 images for the CNN classifications by randomly reflecting these images around the vertical x of the frame and scaling from a factor between 1 and 2 [figure 8]. The assessment of the capability of each pipeline for correctly classifying the images was used as comparison criteria. The classification performance by training the original images was adopted as a baseline. The initial learning rate of the network was adjusted to 0.0001.

imported>Student221: /* References */

2020-11-24T01:14:01Z

References

← Older revision		Revision as of 01:14, 24 November 2020
Line 184:		Line 184:
	[4 ] H. Simon. “Neural Networks: a comprehensive foundation”. Prentice Hall 2nd ed. 1999.		[4 ] H. Simon. “Neural Networks: a comprehensive foundation”. Prentice Hall 2nd ed. 1999.

	[5] ~~K. He, X. Zhang, S.Ren and J. Sun. “Deep Residual Learning for Image Recognition”.~~ https://~~arxiv~~.org/~~abs~~/~~1512.03385~~		[5] https://en.wikipedia.org/wiki/AlexNet

	[6] ~~Resnet-50 Kaggle Pre-trained models~~ for ~~keras:~~ https://~~www.kaggle~~.~~com~~/~~keras~~/~~resnet50~~		[6] K. He, X. Zhang, S.Ren and J. Sun. “Deep Residual Learning for Image Recognition”. https://arxiv.org/abs/1512.03385

	[7] Resnet-50 ~~Matlab~~: https://www.~~mathworks~~.com/~~help/deeplearning/ref~~/resnet50~~.html?s_tid=srchtitle~~		[7] Resnet-50 Kaggle Pre-trained models for keras: https://www.kaggle.com/keras/resnet50

	[8] ~~V. Viswanathan , R. Hussein. “Applications of Image Processing and Real~~-~~Time embedded Systems in Autonomous Cars~~: ~~A Short Review”~~. ~~International Journal of Image Processing (IJIP), Volume (11): Issue (2): 2017~~.		[8] Resnet-50 Matlab: https://www.mathworks.com/help/deeplearning/ref/resnet50.html?s_tid=srchtitle

	[9] H. ~~Zhu~~, K. ~~Yuen, L~~. ~~Mihaylova~~ and H. ~~Leung. “Overview~~ of ~~Environment Perception for Intelligent Vehicles”. IEEE Transactions on Intelligent Transportation Systems, vol.18, No.10~~, ~~October~~ 2017.		[9] V. Viswanathan , R. Hussein. “Applications of Image Processing and Real-Time embedded Systems in Autonomous Cars: A Short Review”. International Journal of Image Processing (IJIP), Volume (11): Issue (2): 2017.

	[10] ~~https://en~~.~~wikipedia~~.~~org/wiki/Defocus_aberration~~		[10] H. Zhu, K. Yuen, L. Mihaylova and H. Leung. “Overview of Environment Perception for Intelligent Vehicles”. IEEE Transactions on Intelligent Transportation Systems, vol.18, No.10, October 2017.

	[11] https://~~digital-photography-school~~.~~com~~/~~chromatic-aberration-what-is-it-and-how-to-avoid-it~~/		[11] https://en.wikipedia.org/wiki/Defocus_aberration

	[12] H. ~~Jiang, Q. Tian, J. Farrell~~ and ~~B. Wandell. “Learning the image processing pipeline”. IEEE Transactions on Image Processing, vol. 26, no. 10, pp. 5032–5042, Oct 2017.~~		[12] https://digital-photography-school.com/chromatic-aberration-what-is-it-and-how-to-avoid-it/

	[13] MLTransferLearning guidelines: https://stanford-pilot.hosted.panopto.com/Panopto/Pages/Viewer.aspx?id=d0f21cb7-7111-43d0-a647-ac4c017dfba0		[13] H. Jiang, Q. Tian, J. Farrell and B. Wandell. “Learning the image processing pipeline”. IEEE Transactions on Image Processing, vol. 26, no. 10, pp. 5032–5042, Oct 2017.

			[14] MLTransferLearning guidelines: https://stanford-pilot.hosted.panopto.com/Panopto/Pages/Viewer.aspx?id=d0f21cb7-7111-43d0-a647-ac4c017dfba0

	== Acknowledgements ==		== Acknowledgements ==

imported>Student221: /* Background */

2020-11-24T01:12:56Z

Background

← Older revision		Revision as of 01:12, 24 November 2020
Line 22:		Line 22:

	'''The modern era of neural networks''' began with the pioneering work of McCulloch and Pitts(1943). They described a logical calculus of neural networks that united the studies of neurophysiology and mathematical logic. With a sufficient number of simple units, neurons, and synaptic connections set properly and operating synchronously, they showed that a network would compute any computable function. The disciplines of neural networks and artificial intelligence were born. The properties of the machines and their behavior are inspired by facts about animal brains. In 1986 the development of the back-propagation algorithm was reported by Rumelhart, Hinton and Williams. Back-propagation learning was discovered independently in two other places about the same time (Parker, 1985; LeCun, 1985). Convolutional Neural Networks allowed a significant improvement in the performance of object detection [4]; and machine vision.		'''The modern era of neural networks''' began with the pioneering work of McCulloch and Pitts(1943). They described a logical calculus of neural networks that united the studies of neurophysiology and mathematical logic. With a sufficient number of simple units, neurons, and synaptic connections set properly and operating synchronously, they showed that a network would compute any computable function. The disciplines of neural networks and artificial intelligence were born. The properties of the machines and their behavior are inspired by facts about animal brains. In 1986 the development of the back-propagation algorithm was reported by Rumelhart, Hinton and Williams. Back-propagation learning was discovered independently in two other places about the same time (Parker, 1985; LeCun, 1985). Convolutional Neural Networks allowed a significant improvement in the performance of object detection [4]; and machine vision.
	AlexNet 2012/Resnet - layer depth enhancement ~~https://en~~.~~wikipedia.org/wiki/AlexNet~~		AlexNet 2012/Resnet - layer depth enhancement[5].

	'''Deep Residual Learning for Image Recognition''', “ResNet-50”, are currently the state-of-the-art in image classification. Deeper neural networks are more difficult to train. Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun presented a residual learning framework to ease the training of networks that are substantially deeper than those used previously. They explicitly reformulated the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. They provided comprehensive empirical evidence showing that these residual networks are easier to optimize, and gain accuracy from considerably increased depth [5,6]. The method was evaluated on the ImageNet 2012 classification dataset that consisted of 1000 classes. The models are trained on the 1.28 million training images and evaluated on the 50k validation images. It was tested on 100k test images.		'''Deep Residual Learning for Image Recognition''', “ResNet-50”, are currently the state-of-the-art in image classification. Deeper neural networks are more difficult to train. Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun presented a residual learning framework to ease the training of networks that are substantially deeper than those used previously. They explicitly reformulated the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. They provided comprehensive empirical evidence showing that these residual networks are easier to optimize, and gain accuracy from considerably increased depth [6,7]. The method was evaluated on the ImageNet 2012 classification dataset that consisted of 1000 classes. The models are trained on the 1.28 million training images and evaluated on the 50k validation images. It was tested on 100k test images.

			'''Pre-trained Model''': A pre-trained model has been previously trained on a dataset and contains the weights and biases that represent the features of whichever dataset it was trained on. Learned features are often transferable to different data. For example, a model trained on a large dataset of animal images will contain learned features like edges or horizontal lines that you would be transferable to our dataset. Pre-trained models are beneficial for many reasons. Using a pre-trained model saves time. Time and compute resources has already been spent to learn a lot of features that the model will likely benefit from [6,7,8].

	'''Pre-trained Model''': A pre-trained model has been previously trained on a dataset and contains the weights and biases that represent the features of whichever dataset it was trained on. Learned features are often transferable to different data. For example, a model trained on a large dataset of animal images will contain learned features like edges or horizontal lines that you would be transferable to our dataset. Pre-trained models are beneficial for many reasons. Using a pre-trained model saves time. Time and compute resources has already been spent to learn a lot of features that the model will likely benefit from [5,6,7].		'''Monochrome images''': Technologies used in autonomous vehicles typically include lane detection. Typical images collected by the on-board camera are color images [9,10]. Each pixel in the image is made up of R, G, and B three color components, which contains large amount of information. Processing these images directly makes the algorithm consume a lot of time. Image preprocessing includes grayscale conversion of color image, gray stretch, median filter to eliminate the image noise and other interference information. Gray stretch can increase the contrast between the lane and the road, which makes the lane lines more prominent. Equation (1) represents the function which is to be applied to an RGB image to convert it to Gray Scale.

	'''Monochrome images''': Technologies used in autonomous vehicles typically include lane detection. Typical images collected by the on-board camera are color images [8,9]. Each pixel in the image is made up of R, G, and B three color components, which contains large amount of information. Processing these images directly makes the algorithm consume a lot of time. Image preprocessing includes grayscale conversion of color image, gray stretch, median filter to eliminate the image noise and other interference information. Gray stretch can increase the contrast between the lane and the road, which makes the lane lines more prominent. Equation (1) represents the function which is to be applied to an RGB image to convert it to Gray Scale.

	L(x,y) = 0.21 R(x,y) + 0.72G(x,y) + 0.07 B(x,y) (1)		L(x,y) = 0.21 R(x,y) + 0.72G(x,y) + 0.07 B(x,y) (1)

	Where R - Red component of the image G - Green component of the image B - Blue component of the image x,y - position of a pixel [8]. In the current study, gray stretch is not performed. However, color images are replaced by monochrome images to evaluating this effect in CNN classification performances.		Where R - Red component of the image G - Green component of the image B - Blue component of the image x,y - position of a pixel [9]. In the current study, gray stretch is not performed. However, color images are replaced by monochrome images to evaluating this effect in CNN classification performances.

	'''Defocus aberration''': In optics, defocus is the aberration in which an image is simply out of focus. Optically, defocus refers to a translation of the focus along the optical axis away from the detection surface. In general, defocus reduces the sharpness and contrast of the image. What should be sharp, high-contrast edges in a scene become gradual transitions. Fine detail in the scene is blurred or even becomes invisible. Nearly all image-forming optical devices incorporate some form of focus adjustment to minimize defocus and maximize image quality [10]. Figure 1 demonstrates typical cases of chromatic aberrations.		'''Defocus aberration''': In optics, defocus is the aberration in which an image is simply out of focus. Optically, defocus refers to a translation of the focus along the optical axis away from the detection surface. In general, defocus reduces the sharpness and contrast of the image. What should be sharp, high-contrast edges in a scene become gradual transitions. Fine detail in the scene is blurred or even becomes invisible. Nearly all image-forming optical devices incorporate some form of focus adjustment to minimize defocus and maximize image quality [11]. Figure 1 demonstrates typical cases of chromatic aberrations.


Line 41:		Line 40:


	Figure 1. Left: intense effect of chromatic aberration (mouth); Center: diagram indicating chromatic aberration produced by the lens; Right: text image shows strong chromatic aberrations [10,11].		Figure 1. Left: intense effect of chromatic aberration (mouth); Center: diagram indicating chromatic aberration produced by the lens; Right: text image shows strong chromatic aberrations [11,12].

	'''Image Signal Processing Pipelines''': Jiang et. al [12] introduced a method that combines machine learning and image systems simulation that automates the image processing pipeline design. The approach is based on a new way of thinking of the image processing pipeline as a large collection of local linear filters. The method has been used to design pipelines for novel sensor architectures in consumer photography applications. It applies a learning-based approach to the ISP pipeline design using affine mapping frameworks. Image patches are clustered based on simple features and then a per-class affine mapping learns to map the raw patches to the sRGB patches. This work combines image systems simulation technology and modern computational methods into a methodology that creates image processing pipelines.”		'''Image Signal Processing Pipelines''': Jiang et. al [13] introduced a method that combines machine learning and image systems simulation that automates the image processing pipeline design. The approach is based on a new way of thinking of the image processing pipeline as a large collection of local linear filters. The method has been used to design pipelines for novel sensor architectures in consumer photography applications. It applies a learning-based approach to the ISP pipeline design using affine mapping frameworks. Image patches are clustered based on simple features and then a per-class affine mapping learns to map the raw patches to the sRGB patches. This work combines image systems simulation technology and modern computational methods into a methodology that creates image processing pipelines.”

	== Methods ==		== Methods ==

imported>Student221: /* Background */

2020-11-24T01:05:11Z

Background

← Older revision		Revision as of 01:05, 24 November 2020
Line 23:		Line 23:
	'''The modern era of neural networks''' began with the pioneering work of McCulloch and Pitts(1943). They described a logical calculus of neural networks that united the studies of neurophysiology and mathematical logic. With a sufficient number of simple units, neurons, and synaptic connections set properly and operating synchronously, they showed that a network would compute any computable function. The disciplines of neural networks and artificial intelligence were born. The properties of the machines and their behavior are inspired by facts about animal brains. In 1986 the development of the back-propagation algorithm was reported by Rumelhart, Hinton and Williams. Back-propagation learning was discovered independently in two other places about the same time (Parker, 1985; LeCun, 1985). Convolutional Neural Networks allowed a significant improvement in the performance of object detection [4]; and machine vision.		'''The modern era of neural networks''' began with the pioneering work of McCulloch and Pitts(1943). They described a logical calculus of neural networks that united the studies of neurophysiology and mathematical logic. With a sufficient number of simple units, neurons, and synaptic connections set properly and operating synchronously, they showed that a network would compute any computable function. The disciplines of neural networks and artificial intelligence were born. The properties of the machines and their behavior are inspired by facts about animal brains. In 1986 the development of the back-propagation algorithm was reported by Rumelhart, Hinton and Williams. Back-propagation learning was discovered independently in two other places about the same time (Parker, 1985; LeCun, 1985). Convolutional Neural Networks allowed a significant improvement in the performance of object detection [4]; and machine vision.
	AlexNet 2012/Resnet - layer depth enhancement https://en.wikipedia.org/wiki/AlexNet		AlexNet 2012/Resnet - layer depth enhancement https://en.wikipedia.org/wiki/AlexNet


	'''Deep Residual Learning for Image Recognition''', “ResNet-50”, are currently the state-of-the-art in image classification. Deeper neural networks are more difficult to train. Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun presented a residual learning framework to ease the training of networks that are substantially deeper than those used previously. They explicitly reformulated the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. They provided comprehensive empirical evidence showing that these residual networks are easier to optimize, and gain accuracy from considerably increased depth [5,6]. The method was evaluated on the ImageNet 2012 classification dataset that consisted of 1000 classes. The models are trained on the 1.28 million training images and evaluated on the 50k validation images. It was tested on 100k test images.		'''Deep Residual Learning for Image Recognition''', “ResNet-50”, are currently the state-of-the-art in image classification. Deeper neural networks are more difficult to train. Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun presented a residual learning framework to ease the training of networks that are substantially deeper than those used previously. They explicitly reformulated the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. They provided comprehensive empirical evidence showing that these residual networks are easier to optimize, and gain accuracy from considerably increased depth [5,6]. The method was evaluated on the ImageNet 2012 classification dataset that consisted of 1000 classes. The models are trained on the 1.28 million training images and evaluated on the 50k validation images. It was tested on 100k test images.

imported>Student221: /* Background */

2020-11-24T01:03:12Z

Background

← Older revision		Revision as of 01:03, 24 November 2020
Line 22:		Line 22:

	'''The modern era of neural networks''' began with the pioneering work of McCulloch and Pitts(1943). They described a logical calculus of neural networks that united the studies of neurophysiology and mathematical logic. With a sufficient number of simple units, neurons, and synaptic connections set properly and operating synchronously, they showed that a network would compute any computable function. The disciplines of neural networks and artificial intelligence were born. The properties of the machines and their behavior are inspired by facts about animal brains. In 1986 the development of the back-propagation algorithm was reported by Rumelhart, Hinton and Williams. Back-propagation learning was discovered independently in two other places about the same time (Parker, 1985; LeCun, 1985). Convolutional Neural Networks allowed a significant improvement in the performance of object detection [4]; and machine vision.		'''The modern era of neural networks''' began with the pioneering work of McCulloch and Pitts(1943). They described a logical calculus of neural networks that united the studies of neurophysiology and mathematical logic. With a sufficient number of simple units, neurons, and synaptic connections set properly and operating synchronously, they showed that a network would compute any computable function. The disciplines of neural networks and artificial intelligence were born. The properties of the machines and their behavior are inspired by facts about animal brains. In 1986 the development of the back-propagation algorithm was reported by Rumelhart, Hinton and Williams. Back-propagation learning was discovered independently in two other places about the same time (Parker, 1985; LeCun, 1985). Convolutional Neural Networks allowed a significant improvement in the performance of object detection [4]; and machine vision.
			AlexNet 2012/Resnet - layer depth enhancement https://en.wikipedia.org/wiki/AlexNet


	'''Deep Residual Learning for Image Recognition''', “ResNet-50”, are currently the state-of-the-art in image classification. Deeper neural networks are more difficult to train. Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun presented a residual learning framework to ease the training of networks that are substantially deeper than those used previously. They explicitly reformulated the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. They provided comprehensive empirical evidence showing that these residual networks are easier to optimize, and gain accuracy from considerably increased depth [5,6]. The method was evaluated on the ImageNet 2012 classification dataset that consisted of 1000 classes. The models are trained on the 1.28 million training images and evaluated on the 50k validation images. It was tested on 100k test images.		'''Deep Residual Learning for Image Recognition''', “ResNet-50”, are currently the state-of-the-art in image classification. Deeper neural networks are more difficult to train. Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun presented a residual learning framework to ease the training of networks that are substantially deeper than those used previously. They explicitly reformulated the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. They provided comprehensive empirical evidence showing that these residual networks are easier to optimize, and gain accuracy from considerably increased depth [5,6]. The method was evaluated on the ImageNet 2012 classification dataset that consisted of 1000 classes. The models are trained on the 1.28 million training images and evaluated on the 50k validation images. It was tested on 100k test images.


	'''Pre-trained Model''': A pre-trained model has been previously trained on a dataset and contains the weights and biases that represent the features of whichever dataset it was trained on. Learned features are often transferable to different data. For example, a model trained on a large dataset of animal images will contain learned features like edges or horizontal lines that you would be transferable to our dataset. Pre-trained models are beneficial for many reasons. Using a pre-trained model saves time. Time and compute resources has already been spent to learn a lot of features that the model will likely benefit from [5,6,7].		'''Pre-trained Model''': A pre-trained model has been previously trained on a dataset and contains the weights and biases that represent the features of whichever dataset it was trained on. Learned features are often transferable to different data. For example, a model trained on a large dataset of animal images will contain learned features like edges or horizontal lines that you would be transferable to our dataset. Pre-trained models are beneficial for many reasons. Using a pre-trained model saves time. Time and compute resources has already been spent to learn a lot of features that the model will likely benefit from [5,6,7].

imported>Student221: /* Slides */

2020-11-24T00:57:17Z

Slides

← Older revision		Revision as of 00:57, 24 November 2020
Line 357:		Line 357:

	[[File:ppt_uar.pdf]]		[[File:ppt_uar.pdf]]


			== Files ==

imported>Student221: /* Slides */

2020-11-24T00:55:13Z

Slides

← Older revision		Revision as of 00:55, 24 November 2020
Line 357:		Line 357:

	[[File:ppt_uar.pdf]]		[[File:ppt_uar.pdf]]


			https://office365stanford-my.sharepoint.com/personal/uarosa_stanford_edu/_layouts/15/onedrive.aspx?id=%2Fpersonal%2Fuarosa%5Fstanford%5Fedu%2FDocuments%2FAttachments%2FDavidCardinal%5FAfricanMammals%5Fscreened%5Ffiles%2Ezip&parent=%2Fpersonal%2Fuarosa%5Fstanford%5Fedu%2FDocuments%2FAttachments&originalPath=aHR0cHM6Ly9vZmZpY2UzNjVzdGFuZm9yZC1teS5zaGFyZXBvaW50LmNvbS86dTovZy9wZXJzb25hbC91YXJvc2Ffc3RhbmZvcmRfZWR1L0VRNTNzNUFnb0poT2thUFI3TldXQjhFQkJMM1VsUzJQY2I1b3JzT2otd3ZWLUE_cnRpbWU9cFFFejd4R1EyRWc

			https://office365stanford-my.sharepoint.com/personal/uarosa_stanford_edu/_layouts/15/onedrive.aspx?id=%2Fpersonal%2Fuarosa%5Fstanford%5Fedu%2FDocuments%2FAttachments%2FDavidCardinal%5FAfricanMammals%5Fprocessed%5Ffiles%2Ezip&parent=%2Fpersonal%2Fuarosa%5Fstanford%5Fedu%2FDocuments%2FAttachments&originalPath=aHR0cHM6Ly9vZmZpY2UzNjVzdGFuZm9yZC1teS5zaGFyZXBvaW50LmNvbS86dTovZy9wZXJzb25hbC91YXJvc2Ffc3RhbmZvcmRfZWR1L0VTV0c3TTVyb0JaQ3ZydHpzeEY5UzMwQk1ra09jdC02TVk5MVQzbW9INGhzd3c_cnRpbWU9RjdHaVZ4S1EyRWc

imported>Student221: /* Slides */

2020-11-24T00:51:02Z

Slides

← Older revision		Revision as of 00:51, 24 November 2020
Line 357:		Line 357:

	[[File:ppt_uar.pdf]]		[[File:ppt_uar.pdf]]


	~~[[Media:https://office365stanford-my.sharepoint.com/:u:/g/personal/uarosa_stanford_edu/EQ53s5AgoJhOkaPR7NWWB8EBBL3UlS2Pcb5orsOj-wvV-A]]~~

imported>Student221: /* Slides */

2020-11-24T00:41:32Z

Slides

← Older revision		Revision as of 00:41, 24 November 2020
Line 359:		Line 359:


	[[Media:~~Example~~.~~ogg~~]]		[[Media:https://office365stanford-my.sharepoint.com/:u:/g/personal/uarosa_stanford_edu/EQ53s5AgoJhOkaPR7NWWB8EBBL3UlS2Pcb5orsOj-wvV-A]]

imported>Student221: /* Slides */

2020-11-24T00:40:35Z

Slides

← Older revision		Revision as of 00:40, 24 November 2020
Line 358:		Line 358:
	[[File:ppt_uar.pdf]]		[[File:ppt_uar.pdf]]

	[[Media:~~https://office365stanford-my.sharepoint~~.~~com/:u:/g/personal/uarosa_stanford_edu/EQ53s5AgoJhOkaPR7NWWB8EBBL3UlS2Pcb5orsOj-wvV-A~~]]
			[[Media:Example.ogg]]