Several types of convolutional neural networks exist, including traditional CNNs, recurrent neural networks, fully convolutional networks and spatial transformer networks — among others.

Traditional CNNs

Traditional CNNs, also known as “vanilla” CNNs, consist of a series of convolutional and pooling layers, followed by one or more fully connected layers. As mentioned, each convolutional layer in this network runs a series of convolutions with a collection of teachable filters to extract features from the input image.

The Lenet-5 architecture, one of the first effective CNNs for handwritten digit recognition, illustrates a conventional CNN. It has two sets of convolutional and pooling layers following two fully connected layers. CNNs’ efficiency in image identification was proved by the Lenet-5 architecture, which also made them more widely used in computer vision tasks.

Recurrent neural networks

Recurrent neural networks (RNNs) are a type of neural network that can process sequential data by keeping track of the context of prior inputs. Recurrent neural networks can handle inputs of varying lengths and produce outputs dependent on the previous inputs, unlike typical feedforward neural networks, which only process input data in a fixed order.

For instance, RNNs can be utilized in NLP activities like text generation or language translation. A recurrent neural network can be trained on pairs of sentences in two different languages to learn to translate between the two. 

The RNN processes sentences one at a time, producing an output sentence depending on the input sentence and the preceding output at each step. The RNN can produce correct translations even for complex texts since it keeps track of past inputs and outputs.

Fully convolutional networks

Fully convolutional networks (FCNs) are a type of neural network architecture commonly used in computer vision tasks such as image segmentation, object detection and image classification. FCNs can be trained end-to-end using backpropagation to categorize or segment images. 

Backpropagation is a training algorithm that computes the gradients of the loss function with respect to the weights of a neural network. A machine learning model’s ability to predict the anticipated output for a given input is measured by a loss function.

FCNs are solely based on convolutional layers, as they do not have any fully connected layers, making them more adaptable and computationally efficient than conventional convolutional neural networks. A network that accepts an input image and outputs the location and classification of objects within the image is an example of an FCN.

Spatial transformer network

A spatial transformer network (STN) is used in computer vision tasks to improve the spatial invariance of the features learned by the network. The ability of a neural network to recognize patterns or objects in an image independent of their geographical location, orientation or scale is known as spatial invariance. 

A network that applies a learned spatial transformation to an input image before processing it further is an example of an STN. The transformation could be used to align objects within the image, correct for perspective distortion or perform other spatial changes to enhance the network’s performance on a specific job.

A transformation refers to any operation that modifies an image in some way, such as rotating, scaling or cropping. Alignment refers to the process of ensuring that objects within an image are centered, oriented or positioned in a consistent and meaningful way. 

When objects in an image appear skewed or deformed due to the angle or distance from which the image was taken, perspective distortion occurs. Applying several mathematical transformations to the image, such as affine transformations, can be used to correct for perspective distortion. Affine transformations preserve parallel lines and ratios of distances between points to correct for perspective distortion or other spatial changes in an image.

Spatial changes refer to any modifications to the spatial structure of an image, such as flipping, rotating or translating the image. These changes can augment the training data or address specific challenges in the task, such as lighting, contrast or background variations.

Source link

Source link

Source link

Source link

In recent times, Bitcoin’s hashrate has been consistently above 300 exahash per second (EH/s) as multiple mining pools dedicate significant hashpower to the Bitcoin blockchain today. Interestingly, some of the world’s top bitcoin mining pools are also using their hashrate to merge-mine other coins, and these networks have benefited from bitcoin’s increased hashrate.

How Bitcoin’s Hashrate Benefits Other Crypto Networks

Bitcoin’s hashrate secures the network and provides rewards for miners participating in the system, but mining pools also dedicate computational power to networks like Namecoin, Elastos, Emercoin, and Vcash. For example, Namecoin has a hashrate of around 187 EH/s today, and some of the top bitcoin mining pools merge-mine the network to acquire namecoin (NMC) rewards.

Merge mining is a process in which miners can mine various cryptocurrencies at the same time without any additional cost. Merged mining is similar to a person playing Pac-Man and Asteroids at the same time, using the same joystick and earning rewards for both games. Namecoin was the first cryptocurrency project to be merge-mined, as it shares the same SHA256 algorithm as Bitcoin, and the first merge-mined block on the network was mined on Sept. 19, 2011.

Bitcoin pools that dedicate hashrate to the Namecoin chain include F2pool, Viabtc, Poolin, and Mining Dutch. While F2pool is the fourth largest bitcoin mining pool over the last three days, it’s the largest namecoin miner as it dedicates its entire 44 EH/s to the Namecoin network. Viabtc dedicates 26.25 EH/s to the Namecoin chain, and Poolin points 5.10 EH/s toward Namecoin as well. At the time of writing, a single namecoin (NMC) is worth $1.24 per unit and 12.5 NMC plus fees are distributed in each block reward.

Namecoin has the second-largest hashrate among SHA256 blockchains, but the Emercoin (EMC) network is the third-largest under BTC and NMC. EMC has 93.38 EH/s dedicated to the network, and Mining Dutch and Viabtc are the top miners for the coin. Viabtc, which is BTC’s fifth-largest mining pool by hashrate, also dedicates 26.76 EH/s to EMC. The Emercoin network leverages a hybrid proof-of-work (PoW) and proof-of-stake (PoS) consensus mechanism. A single emercoin (EMC) is currently changing hands for $0.0088 per coin.

Meanwhile, Viabtc dedicates the same amount of hashrate to the Syscoin (SYS) network, another hybrid PoW and PoS blockchain. Today, a single SYS trades for $0.167 against the U.S. dollar. In addition to the aforementioned PoW cryptocurrencies that leverage the SHA256 consensus algorithm, miners are also dedicating hashrate to networks such as Xaya, Veil, Hathor, Elastos, and Vcash. Older cryptocurrency networks like Terracoin (TRC) and Unobtanium (UNO) also see a small fraction of SHA256 hashrate.

F2pool dedicates 44.32 EH/s to Vcash, but the coin’s native asset has no listed value on any of the top coin market aggregation sites. Elastos has over 100 exahash dedicated to the chain, and top mining pools like Antpool, F2pool, Viabtc, and Mining Dutch are dedicating hashrate to the Elastos network. Current statistics further show that 100 exahash per second is also dedicated to the RSK smart contract network.

What do you think the future holds for merge mining and the relationships between different blockchain networks? Share your thoughts in the comments section below.

Jamie Redman

Jamie Redman is the News Lead at Bitcoin.com News and a financial tech journalist living in Florida. Redman has been an active member of the cryptocurrency community since 2011. He has a passion for Bitcoin, open-source code, and decentralized applications. Since September 2015, Redman has written more than 6,000 articles for Bitcoin.com News about the disruptive protocols emerging today.