gautamhcscv Goto Github PK

ai-ml

attacks-and-defence

Adversarial attacks, their identifications and defenses against them

authentication_system

OTP verifier using python

bias-mitigation-using-knowledge-distillation

A teacher model with minimal bias or calibrate its outputs to reduce bias transfer during distillation. Focus on distilling information from teacher features that are less prone to bias, like semantic representations instead of raw activations.

chest-image-classification

We design a lightweight CNN architecture for the chest x-ray classi-32 fication task by introducing ExLNet which uses a novel DCISE blocks to reduce the33 computational burden. We show the effectiveness of the proposed architecture through34 various experiments performed on publicly available datasets.

chest-radiograph-anomaly-detection

classification-loss-comparison

Densenet121 models is compared with Triple loss, Center loss and Cross-entropy loss. Cross-Entropy Loss Function. Also called logarithmic loss, log loss or logistic loss. Each predicted class probability is compared to the actual class desired output 0 or 1 and a score/loss is calculated that penalizes the probability based on how far it is from the actual expected value. Triplet loss is a loss function for machine learning algorithms where a reference input (called anchor) is compared to a matching input (called positive) and a non-matching input (called negative). Center loss reduces the distance of each data point to its class center. It is not as difficult to train as triplet loss and performance is not based on the selection process of the training data points(triplets). Combining it with a softmax loss, prevents embeddings from collapsing.

cnn_algos_comparison

Comparative Study between different CNN algorithms and comparing them using various evaluation metics.

coursera_assignments

detecting-emerging-technologies-and-their-evolution-using-deep-learning-and-weak-signal-analysis-

We present a multi-layer quantitative approach able to identify future signs from scientific publications on hypersonics by leveraging deep learning and weak signal analysis. The proposed framework can help strategic planners and domain experts better identify and monitor emerging technology trends.

federated-learning-methods-comparison

We utilize the Adversarial Model Perturbations (AMP) regularizer to regularize clients’ models. The AMP regulzaizer is based on perturbing the model parameters so as to get a more generalized model. The claim of AMP regularizer is to reach flat minima and therefore is expected to reach flat minima in FL settings as well.

flask-python

Flask is used for developing web pages.

generative-adversial-networks

A generative adversarial network (GAN) is a machine learning (ML) model in which two neural networks compete with each other to become more accurate in their predictions. GANs typically run unsupervised and use a cooperative zero-sum game framework to learn. Generative adversarial networks (GANs) are algorithmic architectures that use two neural networks, pitting one against the other (thus the “adversarial”) in order to generate new, synthetic instances of data that can pass for real data. They are used widely in image generation, video generation and voice generation.

hfsm-python

Hierarchical state machines are finite state machines whose states themselves can be other state machines. Hierarchy is a useful construct in many modeling formalisms and tools for software design, requirements and testing. We summarize recent work on hierarchical state machines with or without concurrency.

image-anonymization-using-adversarial-attacks

The Fast Gradient Sign Method (FGSM) combines a white box approach with a misclassification goal. It tricks a neural network model into making wrong predictions. We use this technique to anonymize images.

image-classification-on-edges-using-bnn

Classification of images based on edges only. The images are converted to their gray-scale format followed by application of sobel filter to detect the edges. Models are applied on this new set of images with only edges to detect their class.

image_recolor_gans

Grayscale to LAB format and grayscale to RGB format.

language-translation-model

Tensorflow model to translate English text to German language.

machine-learning-labs

modelling-of-inhomogeneous-objects

Modelling viscoelastic deformable objects refers to the process of representing and simulating the behavior of materials that exhibit both viscous and elastic properties when subjected to external forces or deformations.

modelling_viscoelastic_objects

This paper proposes an alternative data-driven hap- tic modeling method of homogeneous deformable objects based on a CatBoost approach – a variant of gradient boosting machine learning approach. In this approach, decision trees are trained sequentially to learn the required mapping function for modeling the objects.

multi-lingual-translation-using-python

number_plate_detection-using-yolo-v7

YOLOv7 is the new state-of-the-art object detector in the YOLO family. According to the paper, it is the fastest and most accurate real-time object detector to date. According to the YOLOv7 paper, the best model scored 56.8% Average Precision (AP), which is the highest among all known object detectors.

object-detection

Object detection using tensorflow using emojis.

prediction-of-sincere-questions

Classify questions as sincere or insincere

scratch-implementation

Implementation of Neural Networks from scratch. Implementing Adam Optimization just by using numpy library, RMSProp, ReLU and it's derivatives are also implemented from scratch. Model is initialized with Xavier weight initializer. Training the model for 25 classes on alphabet-recognition dataset and noticing the trends of different epochs with learning rate.

seq2seq-attention-transformer

A course on transformer model for multi-tasking on text data. Learning about BERT model, T5 model and transformer model for various purposes.

transformers-applications

A transformer is a deep learning model. It is distinguished by its adoption of self-attention, differentially weighting the significance of each part of the input (which includes the recursive output) data. It is used primarily in the fields of natural language processing (NLP) and computer vision (CV).

unitytutorials-finitestatemachines

The code for a short tutorial on finite state machines and how to use them for basic 2D physics-based player movement in Unity/C# (in text or video format).

yolov7_traffic_detecitons

Yolo-v7 model for helmet detection.