The proceedings contain 99 papers. The topics discussed include: fast fragment assemblage using boundary line and surface matching;archaeological fragment reconstruction using curve-matching;profile-based pottery reco...
ISBN:
(纸本)0769519008
The proceedings contain 99 papers. The topics discussed include: fast fragment assemblage using boundary line and surface matching;archaeological fragment reconstruction using curve-matching;profile-based pottery reconstruction;accurately estimating sherd 3D surface geometry with application to pot reconstruction;application of structured illumination in nano-scale vision;noise adaptive channel smoothing of low-dose images;indirect symbolic correlation approach to unsegmented text recognition;background line detection with a stochastic model;estimating tracking sources and sinks;generic event detection in sports video using cinematic features;towards perceptual interface for visualization navigation of large data sets;coevolutionary computation for synthesis of recognition systems;mirror shape recovery from image curves and intrinsic parameters: rotationally symmetric and conic mirrors;parametric subpixel matchpoint recovery with uncertainty estimation: a statistical approach;and tracking random sets of vehicles in terrain.
The proceedings contain 99 papers. The topics discussed include: fast fragment assemblage using boundary line and surface matching;archaeological fragment reconstruction using curve-matching;profile-based pottery reco...
ISBN:
(纸本)0769519008
The proceedings contain 99 papers. The topics discussed include: fast fragment assemblage using boundary line and surface matching;archaeological fragment reconstruction using curve-matching;profile-based pottery reconstruction;accurately estimating sherd 3D surface geometry with application to pot reconstruction;application of structured illumination in nano-scale vision;noise adaptive channel smoothing of low-dose images;indirect symbolic correlation approach to unsegmented text recognition;background line detection with a stochastic model;estimating tracking sources and sinks;generic event detection in sports video using cinematic features;towards perceptual interface for visualization navigation of large data sets;coevolutionary computation for synthesis of recognition systems;mirror shape recovery from image curves and intrinsic parameters: rotationally symmetric and conic mirrors;parametric subpixel matchpoint recovery with uncertainty estimation: a statistical approach;and tracking random sets of vehicles in terrain.
The proceedings contain 99 papers. The topics discussed include: fast fragment assemblage using boundary line and surface matching;archaeological fragment reconstruction using curve-matching;profile-based pottery reco...
ISBN:
(纸本)0769519008
The proceedings contain 99 papers. The topics discussed include: fast fragment assemblage using boundary line and surface matching;archaeological fragment reconstruction using curve-matching;profile-based pottery reconstruction;accurately estimating sherd 3D surface geometry with application to pot reconstruction;application of structured illumination in nano-scale vision;noise adaptive channel smoothing of low-dose images;indirect symbolic correlation approach to unsegmented text recognition;background line detection with a stochastic model;estimating tracking sources and sinks;generic event detection in sports video using cinematic features;towards perceptual interface for visualization navigation of large data sets;coevolutionary computation for synthesis of recognition systems;mirror shape recovery from image curves and intrinsic parameters: rotationally symmetric and conic mirrors;parametric subpixel matchpoint recovery with uncertainty estimation: a statistical approach;and tracking random sets of vehicles in terrain.
The proceedings contain 192 papers. The topics discussed include: feature vector compression based on least error quantization;a comprehensive analysis of deep learning based representation for face recognition;two-st...
ISBN:
(纸本)9781467388504
The proceedings contain 192 papers. The topics discussed include: feature vector compression based on least error quantization;a comprehensive analysis of deep learning based representation for face recognition;two-stream CNNs for gesture-based verification and identification: learning user style;CALIPER: continuous authentication layered with integrated PKI encoding recognition;latent fingerprint image segmentation using fractal dimension features and weighted extreme learning machine ensemble;a comparison of human and automated face verification accuracy on unconstrained image sets;offline signature verification based on bag-of-VisualWords model using KAZE features and weighting schemes;implementation of fixed-length template protection based on homomorphic encryption with application to signature biometrics;seeing the forest from the trees: a holistic approach to near-infrared heterogeneous face recognition;a novel visualization tool for evaluating the accuracy of 3d sensing and reconstruction algorithms for automatic dormant pruning applications;a pointing gesture based egocentric interaction system: dataset, approach and application;multimodal multi-stream deep learning for egocentric activity recognition;sparse kernel machines for discontinuous registration and nonstationary regularization;and accurate small deformation exponential approximant to integrate large velocity fields: application to image registration.
The proceedings contain 280 papers. The topics discussed include: the role of synchronic causal conditions in visual knowledge learning;attention-based natural language person retrieval;Singlets: multi-resolution moti...
ISBN:
(纸本)9781538607336
The proceedings contain 280 papers. The topics discussed include: the role of synchronic causal conditions in visual knowledge learning;attention-based natural language person retrieval;Singlets: multi-resolution motion singularities for soccer video abstraction;hockey action recognition via integrated stacked hourglass network;extraction and classification of diving clips from continuous video footage;accurate and efficient 3D human pose estimation algorithm using single depth images for pose analysis in golf;athlete pose estimation by a global-local network;continuous video to simple signals for swimming stroke detection with convolutional neural networks;application of computervision and vector space model for tactical movement classification in badminton;automatic tactical adjustment in real-time: modeling adversary formations with radon-cumulative distribution transform and canonical correlation analysis;infrared variation optimized deep convolutional neural network for robust automatic ground target recognition;an algorithm for parallel reconstruction of jointly sparse tensors with applications to hyperspectral imaging;deep heterogeneous face recognition networks based on cross-modal distillation and an equitable distance metric;face presentation attack with latex masks in multispectral videos;privacy-preserving understanding of human body orientation for smart meetings;and fast, accurate thin-structure obstacle detection for autonomous mobile robots.
This work presents AFRIFASHION1600, an openly accessible contemporary African fashion image dataset containing 1600 samples labelled into 8 classes representing some African fashion styles. Each sample is coloured and...
详细信息
ISBN:
(纸本)9781665448994
This work presents AFRIFASHION1600, an openly accessible contemporary African fashion image dataset containing 1600 samples labelled into 8 classes representing some African fashion styles. Each sample is coloured and has an image size of 128 x 128. This is a niche dataset that aims to improve visibility, inclusion, and familiarity of African fashion in computervision ***1600 dataset is available here.
In the context of variational auto-encoders, learning disentangled latent variable representations remains a challenging problem. In this abstract, we consider the semi-supervised setting, in which the factors of vari...
详细信息
ISBN:
(纸本)9781665448994
In the context of variational auto-encoders, learning disentangled latent variable representations remains a challenging problem. In this abstract, we consider the semi-supervised setting, in which the factors of variation are labelled for a small fraction of our samples. We examine how the quality of learned representations is affected by the dimension of the unsupervised component of the latent space. We also consider a variational lower bound for the mutual information between the data and the semi-supervised component of the latent space, and analyze its role in the context of disentangled representation learning.
Self-attention is a corner stone for transformer models. However, our analysis shows that self-attention in vision transformer inference is extremely sparse. When applying a sparsity constraint, our experiments on ima...
详细信息
ISBN:
(纸本)9781665448994
Self-attention is a corner stone for transformer models. However, our analysis shows that self-attention in vision transformer inference is extremely sparse. When applying a sparsity constraint, our experiments on image (ImageNet-1K) and video (Kinetics-400) understanding show we can achieve 95% sparsity on the self-attention maps while maintaining the performance drop to be less than 2 points. This motivates us to rethink the role of self-attention in vision transformer models.
The aim of this paper is to demonstrate that a state of the art feature matcher (LoFTR) can be made more robust to rotations by simply replacing the backbone CNN with a steerable CNN which is equivariant to translatio...
详细信息
ISBN:
(纸本)9781665487399
The aim of this paper is to demonstrate that a state of the art feature matcher (LoFTR) can be made more robust to rotations by simply replacing the backbone CNN with a steerable CNN which is equivariant to translations and image rotations. It is experimentally shown that this boost is obtained without reducing performance on ordinary illumination and viewpoint matching sequences.
暂无评论