Trending December 2023 # All You Need To Know About Convolutional Neural Networks! # Suggested January 2024 # Top 13 Popular

You are reading the article All You Need To Know About Convolutional Neural Networks! updated in December 2023 on the website Minhminhbmm.com. We hope that the information we have shared is helpful to you. If you find the content interesting and meaningful, please share it with your friends and continue to follow and support us for the latest updates. Suggested January 2024 All You Need To Know About Convolutional Neural Networks!

This article was published as a part of the Data Science Blogathon

Table of Contents:

What is CNN, Why is it important

Biological Inspiration

Fundamentals of CNN

Various Convolutional Neural Networks

Data Augmentation and Transfer Learning

Code example using Keras

Ever since I learned about CNN, it has become one of my favorite topics in Deep Learning, so in this article, I am going to explain everything related to CNN. “Using this we can provide machines with vision. Now vision, I believe is one of the most important senses that we pose. Sighted people rely on vision for everyday tasks such as navigation, recognize objects, recognize complex human emotions and behavior” these are some of the words of Professor Alexander Amini from ‘MIT Introduction to Deep Learning 6.S191’. Since CNN has been specifically designed for visual tasks (say object recognition) so I can say that CNN is a type of Neural Network that is most often applied to image processing problems.

It can be used to detect and recognize faces, can be used in the medical sector to classify various diseases instantly, another example where CNN can be used is autonomous/self-driving cars.

Biological Inspiration

Photo by MIKHAIL VASILYEV on Unsplash

Let’s try to understand this from the historical perspective first. Hubel and Wiesel, two neuroscientists who got their Nobel prize in medicine mostly focused their research on visual tasks, so in an experiment done in 1959, they anesthetized a cat and inserted a microelectrode into its primary visual cortex. They then project a series of light bars on the screen installed in front of the cat. Interestingly they observed that some neurons showed some activity when this light bar was presented at a specific angle while other neurons got activated to some different angle of light bars. What they found was that there were different neurons for different tasks such as edge detection, motion detection, depth detection, and so on. CNN is inspired by this biological idea and research done on this area.

Having known the basic idea of CNN le us try to understand it in more detail.

Fundamentals of CNN 1. Convolution

As you might have guessed from the name itself convolution convols/merge two functions or information to produce a third function/information.

Let’s consider this gray-scaled image of 6×6 dimensions. Since the machine can only understand binary language, this image would appear to be a 6×6 matrix with all 0s on the left half and all 255 on the right half, since it is a grayscale image. The RHS 3×3 matrix is known as kernel/ filter/ mask/ operator, and the * operator is a convolution operator (here it is Sobel edge detector). We now have a 4×4 matrix after doing component-wise multiplication and addition. When we normalize this matrix, we receive an image with an edge highlighted in it.

2. Padding (p)

What if we want the output matrix to have the same dimensions as the input matrix i.e. (nxn). If you substitute n = 6 from the above formula, you will get an output matrix of size 4×4. If you want the output matrix to be of size 6×6 then it is quite obvious that the input matrix should be of size 8×8, so to change the dimensions of the input matrix, there is a concept of padding. If a pad one extra layer to each side then the initial dimension would be increased by 2 i.e. now n = 8 which will give the output matrix dimension to be 6.

Now the question arises what value should I fill this extra layer with?

There are two approaches to it:

Zero – Padding

Same-value Padding

3. Stride (s)

Basically, the stride is nothing but shifting a kernel matrix over the input matrix by a specific number of cells at a time. It helps to reduce the size of the output matrix.

Finally, the formula combining all these would be: (n x n) * (k x k), padding = p, stride = s ⇒ (((n – k + 2p)

NOTE: Here Convolution in Color Images

So far we have looked at the convolution operation, padding, and stride on a greyscale image, what if the image is a colorful image? A color image can be represented as a 3D tensor since it has 3 matrices of Red, Green, and Blue stacked on top of each other. These RGB matrices are often referred to as 3 channels since most of the concepts in image processing are derived from signal processing in electronics and telecommunications major. Hence we can represent this 3D tensor as NxMxC where C is the no. of channels which is equal to 3 for color images.

Similar to the convolution in 2D matrices, convolution in 3D matrices is also component-wise multiplication followed by addition. One thing to keep in mind here is that the kernel should also have the same number of channels as the input matrix.

Formula: (N x N x C) * (K x K x C) ⇒ (N – K + 1) x (N – K + 1) x 1

NOTE: Convolution of a 3D matrix results in a 2D matrix.

Convolution layer

Unlike image processing where we use predefined kernels like the Sobel edge detector, in CNN we try to learn these kernels.

Here in CNN, we have multiple hyperparameters to play with, like kernel size K, padding p, stride s, and the number of kernels M.

In a real-world scenario, we have a series of convolution layers to train first the low-level features like edges, then mid-level features, then high-level features to get a fairly accurate model.

Max-pooling

It is again a biologically inspired concept that introduces some invariance in the model. For example, the model should be in a position to detect a face in an input image no matter where it is located and no matter what its size is. This is achieved using pooling layers.

This is similar to the kernel we have seen above, here also we can apply the concept of strides, kernel size, etc.

Here in this example, we have a 2 x 2 max-pool kernel with stride as 2. It selects the maximum value from a patch. Another pooling is the mean/average pooling where we get the average value instead of the maximum value.

Various Convolutional Neural Networks:

LeNet, 1998

AlexNet, 2012

VGGNet, 2014

ResNet, 2023

Inception Network, 2023

Before we move on to the code part, it is important that we understand what Data Augmentation and Transfer Learning mean.

Data Augmentation

Since everyone wants their model to be a robust model but creating such a model would require a huge amount of data so Data Augmentation comes to the rescue here. It basically means adding more data to our dataset by rotating, scaling, cropping, flipping, shifting horizontally or vertically, zooming, stretching (shear operation), illumination conditions, etc on images (since CNN) to create more artificial data to train on.

Transfer Learning

The main idea is to use a pre-trained model (which is trained on some dataset ‘X) on a dataset ‘Y’ without training it from scratch. Both in Keras and Tensorflow we have some pre-trained models like VGG16 that is trained on one of the largest object classification datasets ImageNet contained 1000 different categories and the total dataset size is 150GB.

Here we can use transfer learning in various ways like using Bottleneck features (i.e. taking output just at the flattening layer) or by freezing the initial layer that detects only edges, shapes, etc. and changing/learning the last few layers according to the new dataset or use the pre-trained model as the initial model to fine-tune the complete model based on the new dataset.

Let’s Code from __future__ import print_function import keras from keras.datasets import mnist from keras.models import Sequential from keras.layers import Dense, Dropout, Flatten from keras.layers import Conv2D, MaxPooling2D from keras import backend as K batch_size = 128 num_classes = 10 epochs = 12 # input image dimensions img_rows, img_cols = 28, 28 # the data, split between train and test sets (x_train, y_train), (x_test, y_test) = mnist.load_data() if K.image_data_format() == 'channels_first': x_train = x_train.reshape(x_train.shape[0], 1, img_rows, img_cols) x_test = x_test.reshape(x_test.shape[0], 1, img_rows, img_cols) input_shape = (1, img_rows, img_cols) else: x_train = x_train.reshape(x_train.shape[0], img_rows, img_cols, 1) x_test = x_test.reshape(x_test.shape[0], img_rows, img_cols, 1) input_shape = (img_rows, img_cols, 1) x_train = x_train.astype('float32') x_test = x_test.astype('float32') x_train /= 255 x_test /= 255 print('x_train shape:', x_train.shape) print(x_train.shape[0], 'train samples') print(x_test.shape[0], 'test samples') # convert class vectors to binary class matrices y_train = keras.utils.to_categorical(y_train, num_classes) y_test = keras.utils.to_categorical(y_test, num_classes) model = Sequential() model.add(Conv2D(32, kernel_size=(3, 3), activation='relu', input_shape=input_shape)) model.add(Conv2D(64, (3, 3), activation='relu')) model.add(MaxPooling2D(pool_size=(2, 2))) model.add(Dropout(0.25)) model.add(Flatten()) model.add(Dense(128, activation='relu')) model.add(Dropout(0.5)) model.add(Dense(num_classes, activation='softmax')) optimizer=keras.optimizers.Adadelta(), metrics=['accuracy']) model.fit(x_train, y_train, batch_size=batch_size, epochs=epochs, verbose=1, validation_data=(x_test, y_test)) score = model.evaluate(x_test, y_test, verbose=0) print('Test loss:', score[0]) print('Test accuracy:', score[1])

References

The media shown in this article are not owned by Analytics Vidhya and are used at the Author’s discretion.

You're reading All You Need To Know About Convolutional Neural Networks!

All You Need To Know About Reinforcement Learning

How to get into this field?

This is one of the most exciting fields that has garnered attention from so many. For the ones looking to make a career in this, there’s no better time than now to get started.

AWS Deepracer

Signing up for Deepracer will give you an access to a simulator. This further allows the user to select a track, code a reward function, and also adjust tuning parameters. With a default reward function with tuning parameters, one can start training the racer followed by evaluating its performance. AWS Deepracer is undoubtedly a great tool to get started and get acquainted with RL.

Real-world examples

The best way to learn anything is to see how it applies to the real-world. With tons of real-world examples of business, academics, and government organizations experimenting as well as succeeding with this innovative aspect of machine learning, a lot can be learned and understood. Some real-life examples that can help in this aspect are – a grocery store that employs a personalized and recommendation engine that uses reinforcement learning, army deploying vehicles in different parts of the battle area, a robot playing sports, to name a few.

Books

Needless to say, books are certainly one of the best ways to gain knowledge and expertise no matter what the field is. Some of the best books and papers to get going are –

The Basics of Deep Q Learning: With this book, one can surely gain command on Math and processes of Reinforcement Learning.

The Hierarchical Learning paper: This is handy especially for those who want to understand Hierarchical Learning in detail.

Deep RL: As evident as the name, this delves deeper into the subject.

Applications of RL

Some of the applications of RL are in the areas of –

Robotics

Bidding & Advertising

Augmented NLP

Industrial Operations

Supply Chain & Logistics

Traffic Control

Load Balancing

Challenges in RL

Scaling and tweaking the neural network that controls the agent

The agent performs the task as it is and not in the optimal or required way.

Preparing the simulation environment is yet another challenge faced.

Reinforcement learning (RL) is the area of machine learning that is concerned with how software is able to take the right decision. The computer employs a trial and error method in order to find the best possible behavior or path that should be taken in a specific situation. Reinforcement learning is more or less a game-like situation. In order to enable the machine to do what is desired by the programmer, the artificial intelligence gets either rewards or penalties for the actions it performs. Ultimately, the goal is to maximize the rewards. Needless to say, this aspect of machine learning is the most effective way to hint at a machine’s creativity as of now. Consider an example wherein an autonomous vehicle is to be designed with a focus on parameters like minimizing the ride time, reducing pollution, obeying the traffic rules, etc. While coming up with this model, rather than relying on lengthy ‘if-then’ statements, reinforcement learning serves to be a savior. Here, the programmer would focus on preparing the reinforcement learning agent capable enough of learning from the system of rewards and chúng tôi is one of the most exciting fields that has garnered attention from so many. For the ones looking to make a career in this, there’s no better time than now to get started.Signing up for Deepracer will give you an access to a simulator. This further allows the user to select a track, code a reward function, and also adjust tuning parameters. With a default reward function with tuning parameters, one can start training the racer followed by evaluating its performance. AWS Deepracer is undoubtedly a great tool to get started and get acquainted with chúng tôi best way to learn anything is to see how it applies to the real-world. With tons of real-world examples of business, academics, and government organizations experimenting as well as succeeding with this innovative aspect of machine learning, a lot can be learned and understood. Some real-life examples that can help in this aspect are – a grocery store that employs a personalized and recommendation engine that uses reinforcement learning, army deploying vehicles in different parts of the battle area, a robot playing sports, to name a few.Needless to say, books are certainly one of the best ways to gain knowledge and expertise no matter what the field is. Some of the best books and papers to get going are –Some of the applications of RL are in the areas of –The crux of RL is how the agent is trained. Reinforcement learning is, without a doubt, cutting-edge technology and has the potential to transform the world. It is one of those ways that can make a machine creative

All You Need To Know About Apple’S New Programming Language

During WWDC 2014, Apple introduced a new programming language called Swift. Swift is intended to be a modern and easy to use language that allows for quicker app development and avoids complexities in XCode.

Apple has simplified programming with its developer tools and extensive documentation over the past few years, but Objective-C and XCode still require a steep learning curve. This can really be uninviting to those who have other options for app development. Even if you’re a professional developer, with seasoned data manipulation skills in higher-level programming languages, you might find yourself annoyed by some of XCode’s nuisances, especially if all you want to do is give your logic a practical shape, without having to manage the difficult and tiresome syntax involved.

Too Much Syntax in Obj-C

Apple’s intent with Swift is to do away with the worrisome syntax, and instead get down to the logic of programming. The language, according to Apple, is built to be safe and manage memory automatically. It also fully supports unicode, so you can store values in standard English variable names, Chinese characters and emoticons if you like.

Integration With Obj-C

For all of you who already know Objective C: You’re in luck. Swift is built to work with the already existing Objective-C programming language, so it should integrate well with all your current Objective-C projects. Swift brings up code of OS X and iOS to a slightly higher level, allowing people with coding experience to develop programs much, much easier.

Apple says that it created the code after extensive research regarding what developers like and don’t like about current programming language.

Swift is the modern programming language for the modern developer – Apple

Swift’s Main New Feature – Playgrounds

Swift’s main new feature that has many developers buzzing is “Playgrounds,” which provides users with live feedback as they code. Resultantly, this makes it possible to test exactly what is going to happen inside an application, without having to compile the whole application. Playgrounds also give complete control of time inside a program so users can see what their code is doing moment by moment. This can arguably save valuable minutes for developers and will surely be a feature that most developers would like to have and use.

Designed For Safety And Easiness

Apple states that Swift is primarily designed for safety and for improving memory corruption bugs. Many developers who have experimented with the new language have stated that it’s much easier to read because of the fact that parameters are expressed in a cleaner syntax; that makes projects in Swift easier to maintain and read. For example, here’s a simple “Hello, World” program in both Objective C and Swift:

In Objective C:

In Swift:

println

(

"Hello, world"

)

It’s that simple.

Swift will additionally help to catch coding errors before they can make it into the final product. This should really help increase developer productivity and help make more stable apps. Developers will be able to submit OS X and iOS apps made with Swift to the App Store when the new operating systems release this fall.

Apple has released an iBooks guidebook for developers using Swift, which is available for free. The book will really help new developers learn Swift, so if you’re an aspiring developer, do check it out.

On June 1st, nobody outside of Apple had heard of Swift. Twenty-fours hours later, it’s a completely different story. Tens of thousands of developers were thinking and planning what they planned to do with it.

It’s true that change comes quickly, rather swiftly these days.

Shujaa Imran

Shujaa Imran is MakeTechEasier’s resident Mac tutorial writer. He’s currently training to follow his other passion become a commercial pilot. You can check his content out on Youtube

Subscribe to our newsletter!

Our latest tutorials delivered straight to your inbox

By signing up, you agree to our Privacy Policy and European users agree to the data transfer policy. We will not share your data and you can unsubscribe at any time.

All You Need To Know About Logistics Technology Trends In 2023

Technologies have redefined the logistics industry and offering strategic methods to manage sales order, inventory, payment, and deliver the orders. However, emerging technologies such as Big Data, Blockchain, and Virtual Reality have enhanced communication, data analysis, and infrastructure, which helps logistics companies to stay ahead in the functional areas of management.

Only opting for these new innovations will not lead the entities towards success. In order to boost the overall business efficiency and performance, it is essential to select the right technology as per the organizational infrastructure and policies.

Advantages of Logistic Solutions Real-Time Visibility

Increase Customer Satisfaction Saves Time 24*7 Monitoring

With logistic software, employees can monitor the system from across the boundaries also. They can monitor the work process for 24*7 hours with smart devices, it is easy to track the vehicles and obtain real-time information.

Automated Functionality

Automation is the major benefit of logistic software. It helps in the planning involved prior to the dispatch. The logistics solution helps to plan the shipment routes and manage the load. Along with this, it excludes the paperwork and turns the process in an effective and timely manner.

Features to be integrated into the Logistic Solution

Everything You Need To Know About Uranium

Since the German chemist Martin Heinrich Klaproth identified uranium in 1789, atomic number 92 has become one of the most troubling substances on the planet. It’s naturally radioactive, but its isotope uranium-235 also happens to be fissile, as Nazi nuclear chemists learned in 1938, when they did the impossible and split a uranium nucleus in two. American physicists at U.C. Berkeley were soon to discover they could force uranium-238 to decay into plutonium-239; the substance has since been used in weapons and power plants around the world. Today, the element continues to stoke international tensions as Iran stockpiles uranium in defiance of an earlier treaty, and North Korea’s “Rocket Man” leader Kim Jong-un continues to resist denuclearization.

But what is uranium, exactly? And what do you need to know about it beyond the red-hot headlines? Here we answer your most pressing nuclear questions:

Where does uranium come from?

Uranium is a common metal. “It can be found in minute quantities in most rocks, soils, and waters,” geologist Dana Ulmer-Scholle writes in an explainer from the New Mexico Bureau of Geology and Mineral Resources. But finding richer deposits—the ones with concentrated uranium actually worth mining—is more difficult.

When engineers find a promising seam, they mine the uranium ore. “It’s not people with pickaxes anymore,” says Jerry Peterson, a physicist at the University of Colorado, Boulder. These days, it comes from leaching, which Peterson describes as pouring “basically PepsiCola—slightly acidic” down into the ground and pumping the liquid up from adjacent holes. As the fluid percolates through the deposit, it separates out the uranium for harvesting.

Uranium ore. Deposit Photos

What are the different types of uranium?

Uranium has several important isotopes—different flavors of the same substance that vary only in their neutron count (also called atomic mass). The most common is uranium-238, which accounts for 99 percent of the element’s presence on Earth. The least common isotope is uranium-234, which forms as uranium-238 decays. Neither of these products are fissile, meaning their atoms don’t easily split, so they can’t sustain a nuclear chain reaction.

That’s what makes the isotope uranium-235 so special—it’s fissile, so with a bit of finessing, it can support a nuclear chain reaction, making it ideal for nuclear power plants and weapons manufacturing. But more on that later.

There’s also uranium-233. It’s another fissile product, but its origins are totally different. It’s a product of thorium, a metallic chemical much more abundant than uranium. If nuclear physicists expose thorium-232 to neutrons, the thorium is liable to absorb a neutron, causing the material to decay into uranium-233.

Just as you can turn thorium into uranium, you can turn uranium into plutonium. Even the process is similar: Expose abundant uranium-238 to neutrons, and it will absorb one, eventually causing it to decay to plutonium-239, another fissile substance that’s been used to create nuclear energy and weapons. Whereas uranium is abundant in nature, plutonium is really only seen in the lab, though it can occur naturally alongside uranium.

How do you go from a rock to a nuclear fuel source?

People don’t exactly lay out step-by-step guides to refining nuclear materials. But Peterson got pretty close. After you’ve extracted uranium from the earth, he says chemical engineers separate the uranium-rich liquid from other minerals in the sample. When the resulting uranium oxide dries, it’s the color of semolina flour, hence the nickname “yellowcake” for this intermediate product.

From there, a plant can purchase a pound of yellowcake for $20 or $30. They mix the powder with hydrofluoric acid. The resulting gas is spun in a centrifuge to separate from uranium-238 and uranium-235. This process is called “enrichment.” Instead of the natural concentration of 0.7 percent, nuclear power plants want a product that’s enriched to between 3 and 5 percent uranium-235. For a weapon, you need much more: These days, upwards of 90 percent is the goal.

Once that uranium is enriched, power plant operators pair it with a moderator, like water, that slows down the neutrons in the uranium. This increases the probability of a consistent chain reaction. When your reaction is finally underway, each individual neutron will transform into 2.4 neutrons, and so on, creating energy all the while.

Uranium glass dinnerware. Deposit Photos

Any fun facts I should take with me to my next dinner party?

Try this: In PopSci‘s “Danger” issue earlier this year, David Meier, a research scientist at Pacific Northwest National Lab, talked about his work to create a database of plutonium sources. Turns out, every plutonium product has a visible origin story, because “there’s not one way of processing it,” Meier says. The United States had two plutonium production sites. While the intermediate product from Hanford, Washington (the Manhattan Project site from which PNNL grew) was brown and yellow, the Savannah River site in Akon, South Carolina, produced “a nice blue material,” Meier says. Law enforcement officials hope these subtle differences—which may also correspond to changes in the chemical signature, particle size, or shape of the material—will one day help them track down illicit nuclear development.

Or, dazzle your guests with a short history of radioactive dinnerware. The manufacture of uranium glass, also called canary glass or Vaseline glass began in the 1830s. Before William Henry Perkin created the first synthetic color in 1856, dyes were terribly expensive and even then they didn’t last. Uranium became a popular way to give plates, vases, and glasses a deep yellow or minty green tinge. But put these household objects under a UV light and they all fluoresce a shocking neon chartreuse. Fortunately for the avid collectors who actively trade in uranium glass, most of these objects aren’t so radioactive as to pose a risk to human health.

Last one: In 2002, the medical journal The Lancet published an article on the concerning potential for depleted uranium—the waste leftover after uranium-235 extraction—to end up on the battlefield. The concern is that its high density would make it an incredible projectile, capable of piercing even the most well-enforced battle tank. Worse yet, it could then contaminate the surrounding landscape and anyone it.

Motorola Moto Z3: All You Need To Know

Motorola held a press event at its Chicago HQ on August 2 and out came the Moto Z3, the company’s 2023 flagship phone but with 2023-era specs. And no, this time there’s no Force variant, just like there was no standard variant in 2023.

As you may have guessed, the Motorola Moto Z3 is superior to the Moto Z3 Play that came out earlier and just like the Moto Z handsets before it, you can only get the Z3 in the U.S. via Verizon Wireless, the exclusive carrier.

Moto Z3 specs

6.01-inch 18:9 FHD+ AMOLED display

Qualcomm Snapdragon 835 processor

4GB RAM

64GB expandable storage

Dual 12 + 12MP main camera

8MP front camera

3000mAh battery

Android 8.1 Oreo

Extras: Bluetooth 5.0, USB-C, NFC, Face unlock, side-mounted scanner, fast charging, water-repellant coating, Moto Display, Voice, Actions, etc.

A closer look at the Moto Z3 reveals quite some resemblance with the Moto Z3 Play. The design language is largely the same because, Moto Mods. Speaking of which, the Z3 was accompanied by a Moto Mod that adds 5G support to the device, which is pretty cool to have.

The two also share some specs such as the display screen, side-mounted fingerprint scanner, face unlock, missing 3.5mm audio jack in favor of a USB-C port, and dual-lens cameras on the back, although the Moto Z3’s setup is obviously superior. However, under the hood, this is easily a 2023 smartphone.

Related: Motorola Moto Z3 and Moto Z3 Play software update news

2023 specs in a 2023 body

The Motorola Moto Z3 is an imitation of what LG has been doing in the recent past. The phone packs 2023 specs in a 2023 body and even though not necessarily a good thing, it has been reflected in the modest asking price for the device.

What this means is that you are not getting a Snapdragon 845, instead, the Moto Z3 has a Snapdragon 835 chipset. Unlike most 2023 flagship phones that have 6GB/8GB RAM modules and 128GB or more storage, the Z3 only gets 4GB of RAM and 64GB of storage, but it can be expanded up to 2TB. Note that the Moto Z2 Force got a 6GB RAM variant with 128GB of internal storage, things you won’t find on the Moto Z3. As noted earlier, you get the same display screen properties as the Moto Z3 Play, but the battery is a little bigger than the mediocre 2730mAh unit found in the Moto Z2 Force.

Like most phones in 2023, the Moto Z3 has a dual-lens camera on the back. Of course, the setup is better than the Play variant, especially the secondary lens that offers black and white shots and is expected to improve performance in low-light conditions. If anything, the dual 12MP setup on the Z3 is identical to the one used on the Moto Z2 Force and for the selfies, you get an 8MP unit.

In 2023, display screens on flagship phones are protected by the latest from Corning – Gorilla Glass 5, but on the Moto Z3 you’ll come across Gorilla Glass 3. Like the Moto Z3 Play and pre-2023 Sony Xperia flagship phones, the fingerprint scanner is side-mounted. The phone also supports facial recognition technology.

Moto Z3 price and availability

The Moto Z3 was unveiled on the second day of August 2023 and starts shipping on August 16th. As pointed out earlier, the phone will be sold through one U.S. carrier – Verizon Wireless. Given the fact that the Z3 has 2023 specs in a 2023 body, the price tag much reflects the former than the latter.

Related: The best Verizon phones to buy today

Verizon says you only need $20 per month for 24 months to grab the Moto Z3 off their shelves. If paying it outright, you only need $480. This is even cheaper than the Moto Z3 Play, which has a price tag of $500, making the Z3 one of the best deals you can get right now.

Motorola’s 5G Mod makes the Z3 future-proof, with the accessory expected to start shipping in early 2023. At the moment, there are no price details, but Motorola says it will be ready at the same time Verizon’s 5G network fully goes mainstream.

Update the detailed information about All You Need To Know About Convolutional Neural Networks! on the Minhminhbmm.com website. We hope the article's content will meet your needs, and we will regularly update the information to provide you with the fastest and most accurate information. Have a great day!