After completing Course #4 of the Coursera Deep Learning specialization I wanted to write a short summary to help y’all understand / brush up on the concept of Convolutional Neural Network (CNN). Let’s understand CNNs with an example –  Figure 1. CNN Example – Source: Coursera DL Specialization Let’s say you have a 32×32 image of digits from 0 to 10 with 3 channels (RGB). You pass it through a filter of size f in the 1st Convolutional Layer (CL1).  What is the size of the output image of the filter? The size of the output image is calculated by the following formula: Source: Medium In our case, let’s assume padding is 0 and stride is 1.  The above formula results in output size of 28×28 for both the height and width of the image. Alright that’s a good start! Let’s keep going.  Notice the dimension 6 on the output of Layer 1.  Where do we get the third dimension from? The third dimension is nothing but the number of filters in the layer. Given a filter of size f. There are #f number of filters in the layer and the dimension of EACH filter is of dimension f x f...

The use of ChatGPT and other AI language models has skyrocketed. Everyone from marketers to website owners, students, and researchers are using them. However, there’s a big question regarding the content generated by these tools.  Is it AI-plagiarism free? While you might be hoping yes, the actual answer is a resounding “No!” AI-generated content commits two forms of plagiarism: (1) typical plagiarism and (2) AI plagiarism.  This article will run through each one, enlightening you about the importance of checking ChatGPT-generated content for any forms of copying. It will also show you the best tools to check for plagiarism in content created by ChatGPT. Without further ado, let’s dive in. Do You Really Need To Check ChatGPT Content for Plagiarism? We can’t stress this enough: whoever you are, if you use ChatGPT for content creation, you must check the output for plagiarism.  Here are four reasons why it really does matter. 1. So you can ensure originality Academics don’t consider ChatGPT content as original. Stanford University has gone as far as to name it “CheatGPT,” with an article on the university’s website stating the use of ChatGPT for content creation ‘clearly violates academic integrity.’ The university also believes it doesn’t...

The healthcare industry has always been at the forefront of technological innovation, from the invention of pacemakers and X-rays to the adoption of electronic health records. Now, Artificial Intelligence (AI) and its allied technologies, such as machine learning, deep learning, and generative AI, are driving the next wave of transformation. Generative AI, in particular, is emerging as a powerful tool with the potential to revolutionize how healthcare is delivered, managed, and experienced. The Rise of Generative AI in Healthcare Generative AI refers to a class of AI models that can generate new, realistic data instances that resemble their training data. Unlike traditional AI, which focuses on analyzing and predicting outcomes, generative AI can create novel content, such as images, text, and even synthetic data. In healthcare, generative AI is being applied to a wide range of use cases, from drug discovery and personalized medicine to medical imaging and patient care. It enhances traditional machine learning applications and opens new possibilities for innovation. Key Benefits of Generative AI in Healthcare Accelerated Drug Discovery: One of the most promising applications of generative AI is in drug discovery. Traditional drug development is a lengthy and expensive process, often taking years and costing billions...

Traditional drone navigation systems rely on pre-mapped environments or slow real-time calculations, limiting their ability to react dynamically. However, researchers from the University of Hong Kong have developed the safety-assured high-speed aerial robot (SUPER) – a compact MAV with a 280-millimeter wheelbase and a thrust-to-weight ratio greater than 5.0. This next-generation autonomous drone can make split-second decisions using LiDAR-based perception, real-time mapping, and a two-trajectory strategy. Its advanced navigation system allows SUPER to navigate complex environments at high speeds while avoiding obstacles in real time. By integrating LiDAR sensing and an intelligent planning framework, the drone ensures both agility and safety – even in completely unknown terrain. The key innovation behind SUPER is its two-trajectory planning system, which allows it to explore new paths while always maintaining a backup plan for safety: Exploratory Trajectory – Charts a fast, efficient path toward its goal, even through unknown spaces. Backup Trajectory – Ensures the drone can always return to a known safe space if the exploratory path encounters obstacles. This system recalculates trajectories 10 times per second, allowing SUPER to react instantly to changes in its surroundings and avoid collisions – even at high speeds. If a replan fails, SUPER will execute...

The model is composed of three sub-networks: Given $x$ (image), encode it into a distribution over the latent space – referred to as $Q(z|x)$ in the previous post. Given $z$ in latent space (code representation of an image), decode it into the image it represents – referred to as $f(z)$ in the previous post. Given $x$, classify its digit by mapping it to a layer of size 10 where the i’th value contains the probability of the i’th digit. The first two sub-networks are the vanilla VAE framework. The third one is used as an auxiliary task, which will enforce some of the latent dimensions to encode the digit found in an image. Let me explain the motivation: in the previous post I explained that we don’t care what information each dimension of the latent space holds. The model can learn to encode whatever information it finds valuable for its task. Since we’re familiar with the dataset, we know the digit type should be important. We want to help the model by providing it with this information. Moreover, we’ll use this information to generate images conditioned on the digit type, as I’ll explain later. Given the digit type, we’ll encode...

Once again a legal tech company has – via the launch of a new AI product – raised the question: has the time come for Small Law to transform itself – and perhaps do so faster than Big Law? But first, the product, which is priced to be affordable to smaller firms. Gavel, which is based in Los Angeles, has built Exec, a genAI assistant that can be embedded in Microsoft Word. It’s already been tried out by several law firms and they’ve provided feedback to ‘ensure accuracy and quality of the results and helped develop market benchmarks and playbooks, starting with corporate and real estate law’, the company noted. All well and good, but what does it do? Gavel Exec ‘empowers lawyers to perform a range of activities, including contract analysis and redlining, negotiation based on firm precedents, and running playbooks with pre-defined rules’, they said, and added that it acts as an ‘extension of Gavel’s end-to-end automation platform’. And here’s several of its features: – Redline: Compare your document against Gavel’s benchmarks or your uploaded reference files to ensure consistency and alignment with your preferred rules and style. – Draft and Revise: Draft or rewrite clauses or entire...