Introduction to Capsule Networks for Computer Vision
Explore Capsule Networks for computer vision, covering sparse pattern recognition, viewpoint equivariance, object detection, and dynamic routing techniques.
Iterative Projection and Matching: Finding Structure-Preserving Representatives and Applications
Explore iterative projection and matching for finding structure-preserving representatives, with applications in active learning, GANs, image processing, and video summarization.
Learning to See in the Dark - Extreme Low Light Imaging Techniques
Explore extreme low-light imaging techniques using deep learning, covering dataset creation, network architecture, and experimental results for improved night vision photography.
Describing Images Using Semantic Modeling with Attributes and Tags
Explore semantic modeling for image description using attributes and tags, covering frameworks, methodologies, and experiments in visual content analysis and popularity prediction.
AttnGAN: Fine-Grained Text to Image Generation with Attentional Generative Adversarial Networks
Explore fine-grained text-to-image generation using attentional GANs, covering architecture, experiments, and novel scenarios in this technical deep dive.
Video Action Transformer Network: Implementation and Experiments
Learn about Video Action Transformer Networks, exploring their architecture, implementation, and experimental results for advanced video understanding and action recognition.
Discovery of Latent 3D Keypoints via End-to-End Geometric Reasoning
Explore end-to-end geometric reasoning for latent 3D keypoint discovery, combining keypoint and geometry learning networks to outperform supervised methods in semi-supervised keypoint finding.
Deep Network for Integrated 3D Sensing of Multiple People in Natural Images
Explore advanced techniques for 3D human pose estimation and shape reconstruction in natural images using deep learning and computer vision approaches.
Learning to Navigate in Cities Without a Map
Explore reinforcement learning techniques for urban navigation without maps, focusing on value functions, courier tasks, and multi-site experiments.
Taskonomy - Disentangling Task Transfer Learning
Explore task transfer learning, higher-order transfers, and task-specific networks. Discover global taxonomies, evaluate computed structures, and assess generalization to novel tasks.
VideoCapsuleNet - A Simplified Network for Action Detection
Explore VideoCapsuleNet, a simplified network for action detection in videos. Learn about capsule networks, convolutional layers, and innovative techniques for improved video analysis.
Neural-Symbolic VQA - Disentangling Reasoning from Vision and Language Understanding
Explore neural-symbolic visual question answering, disentangling reasoning from vision and language understanding through task breakdown, architecture overview, and practical applications.
Adaptive Attacks to Adversarial Example Defenses - CAP6412 Spring 2021
Explore adaptive attacks on adversarial example defenses, covering expectation over transformation, K winners take all, chaos partitioning, and local gradient estimation.
Improving Transferability of Adversarial Examples with Input Diversity - CAP6412 Spring 2021
Explore techniques to enhance adversarial example transferability using input diversity, including FGSM family methods and ensemble network attacks.
Universal Adversarial Perturbations in Machine Learning - Spring 2021
Explore universal adversarial perturbations in machine learning, covering definitions, contributions, experiments, and cross-model universality.
Explaining and Harnessing Adversarial Examples in Machine Learning - Spring 2021
Explore adversarial examples in machine learning, their impact on linear and non-linear models, and techniques like adversarial training to improve model robustness and generalization.
Intriguing Properties of Neural Networks - Spring 2021
Explore intriguing properties of neural networks, including adversarial attacks, transfer, training overlap, and semantic information in units. Analyze weaknesses and spectral instability.
Introduction to Adversarial Attacks in Machine Learning - Lecture 1
Introduction to adversarial attacks on AI systems, covering real-world examples, attack methods, and defense strategies for various applications like face recognition and object detection.
Fast Is Better Than Free: Revisiting Adversarial Training - CAP6412 Spring 2021
Explore adversarial training techniques, comparing PGD with newer methods. Learn about technical contributions, experimental results, and implications for machine learning security.
Explainable AI: Interpreting, Explaining and Visualizing Deep Learning
Explore techniques for interpreting and explaining deep learning models, focusing on LRP rules, Taylor decomposition, and properties of effective explanations in AI systems.
CAP6412 - SmoothGrad: Removing Noise by Adding Noise - Lecture
Explore SmoothGrad, a technique for enhancing neural network visualization by adding noise to input images and averaging resulting sensitivity maps.