Artificial Intelligence

Apple Deep Fusion is a major feature in a camera-phone designed for the consumption of the masses. Similar techniques have been used in the past to get better results from the high end cameras [1] to stay relevant with the advent of better camera functions in people's smartphones, which are able to rival them with their relative ease of use as well as the quality. Now the scenario is possibly going to change, with the onset of Apple’s Deep Fusion . In the words of Apple’s Phil Schiller “The computational photography is a Mad Science” [5] . Phil Schiller presenting latest features in iPhone 11 How it works Apple has included three cameras [2] (visual in the image below) in their latest device that will, each one generating multiple images with variable aperture controls in order to maximize the capturing of the required attributes. What it effectively comes on to is that when you click your device to capture a frame in the specific mode, your camera will actually capture four dif...

MS Vision AI Kit vs Deep Lens Introduction Artificial intelligence has glorified many sections of recent times and among them, one of the most notable ones is the computer vision . Computer vision-based AI can have significant contributions in the classification of specific images, object detection, and segmentation and all of that can be used directly or indirectly in consumer aspects. From face recognition in verifying user identity to digital number plate detection in vehicles, computer vision is playing a significant role and undeniably gaining much popularity. Algorithms Neural networks and some specific deep learning algorithms are the most important constituents of the computer vision. Among those algorithms, CNN 's, GAN 's, and SSD 's are notable. For a newbie or even for an average skilled machine learning developer the Convolutional Neural Network (CNN) is one of the most complex sections. CNNs are not only complex to understand but also they have some spec...

Quantum computing is a sub field quantum information science in which has its focus on developing computer technology from the learnings of quantum theory [1] , by trying to leverage the quantum phenomena like entanglement [2] , super-position [3] to enhance the computational capacity of a quantum computer. Quantum computing is able to produce great results in terms of its computational capabilities and the sheer time difference in performance of complex operations is baffling. The major aspect that causes such a huge variation in the performance with respect to the traditional computing techniques is the fundamental building block difference in between the two. Traditional computers have bits [4] at their foundations that can be only take binary values, i.e. they can be either 1 or they can be 0. This rigidity as is evident only allows for two discrete values to be possible at a given time, therefore any complex higher order representation has to be formed with the combination of t...

Human machine interaction has always been a hot topic, specifically after the advent of Artificial intelligence. The social impact of such interactions is huge to say the very least. AI has more often than not has taken inspiration from the biological neural-architectures to be able enough, to do a significant number of tasks, tasks at which, it is now able to outperform the biological neurons as well [1] . There have been a number of attempts [2] to utilise AI to enhance human capabilities through Brain-machine-interfaces or BMIs [3] . Historically the integration of the BMIs into the human skull has been quite an invasive process. The normal procedures generally required a complete surgery, where the skull was cut open and the electrodes inserted into the brain tissue with the opening then sewed back together. The procedure was not only invasive and complicated in nature but also ran the risk of catching an infection or can be the catalyst for a stroke for the individual receiving...

What is a GAN? GANs or Generative Adversarial Networks were invented by Ian Goodfellow [1] in the year 2014. In a GAN architecture two neural networks are in an adversarial role (therefore the name adversarial network), where we have a generator and a detector or a discriminator. The generator attempts to create a dataset such that the discriminator isn't able to detect if the data is real (from any given distribution) or it is fake (generated by the generator). The generator is given a random initial input or noise from which it learns to generate the data, this generated data is then fed into the discriminator, where we simultaneously input real-data. The discriminator here generates a label (in reality it is a probability distribution) for the fake (or the generated data from generator) and the real dataset. A basic flow can be seen in the following figure : GAN Arch. Source: [2] How Does it Really Learn ? For any type of learning to take place we need to decide an objective ...

Instance Segmentation [1] is one of the hottest areas for research in the field of Artificial Intelligence . The process in its present form relatively computationally expensive, therefore the present architectures that are able to generate the best results are not very suitable for “real-time” applications. The networks that yield good results in instance segmentation are FCIS [2] , Mask-R CNN [3] , RetinaMask [4] , PA-Net [5] etc. These frameworks although perform relatively well but the inference obtained from them can’t be used in “real-time” due to the computational complexity that is involved in the creation of such systems, the sheer number of parameters makes it impossible for these network to perform on machines with lower computational capability. The task therefore requires a different architecture that is able to perform in the “real time” computations. YOLACT to the rescue YOLACT [6] (You only look at the coefficients) is a more optimized version for instance segmentati...

With the advancement in modern technologies, Artificial Intelligence (AI) has made its presence felt in the market. Object detection technology is a hot topic in today’s scenario. Most big companies are making great use of face detection, still image object counting, amongst others. We all know how efficiently computer vision object detection models run on desktop and cloud services. However, in some cases these AI models would require small size devices or hardware for a mobile user. It would also provide the user with the much sort-after aspect of privacy and this is the reason why TensorFlow Lite (TF Lite) came into existence. Detecting Living and Non- Living Objects What Exactly is TensorFlow Lite? TensorFlow Lite is a technology specially designed for mobile phones and smart devices by TensorFlow. As the name suggests, 'Lite' stands for lightweight. It is highly advantageous when looking at the latest technological scenario. TensorFlow allows running machine-learned mo...