This research reverses feature visualization in neural networks to optimize for specific feature objectives.
problem The invertibility of feature visualization in neural networks is not well understood.
method The approach involves optimizing for the feature objective that generates the input used in feature visualization, using the gradient of a specific objective function.
result A closed-form solution is found to minimize the gradient, providing an alternative view on network sensitivity.
Methodology for visualizing labeled datasets with mixed features.
problem Visualization of labeled mixed-featured datasets.
method Developed a Max-Ratio Projection (MRP) method for continuous features and extended it to datasets with discrete and continuous features using Gaussianized distributional transforms and copula models.
result Visualization of labeled mixed-featured datasets using Max-Ratio Projection and Gaussianized distributional transforms.
Proposes a method for automatic feature identification and visual explanation of deep models.
problem Improving interpretability and explanation of deep neural networks.
method Automatic feature identification and visualization of relevant features for classes without additional annotations.
result Produces detailed explanations with good coverage of relevant features.
New method visualizes tabular feature semantics for better model understanding.
problem Lack of feature interaction interpretation in tabular ML models.
method Feature Vectors method for global tabular dataset interpretability.
result Visualizes semantic relationships among tabular features.
GraphTSNE visualizes graph data by integrating graph structure and node features.
problem Lack of suitable visualization techniques for graph-structured data.
method GraphTSNE combines t-SNE with graph convolutional networks to visualize graph data.
result GraphTSNE produces better visualizations of graph data compared to existing methods.
WAPPO optimizes feature distributions for better visual transfer in RL.
problem Improving visual transfer in reinforcement learning.
method WAPPO uses Wasserstein Confusion to minimize feature distribution distance.
result WAPPO outperforms previous methods in visual transfer across different environments.
FeatureEnVi aids in feature engineering with visual analytics.
problem Insufficient support for feature engineering in visual analytics tools.
method Stepwise selection and semi-automatic extraction approaches.
result Extracts heavily engineered features evaluated by multiple metrics.
New visual tools show feature importance for black box models.
problem Improving transparency and trust in machine learning models.
method Local feature importance, PI and ICI plots, partial dependence, individual conditional expectation.
result Visual tools accurately represent feature importance for black box models.
FiLM layers improve visual reasoning tasks by modulating features.
problem Visual reasoning tasks that require multi-step, high-level processes.
method General-purpose FiLM layers that apply feature-wise linear transformations based on conditioning information.
result FiLM layers reduce error by half on the CLEVR benchmark and improve feature coherence.
DEN creates interpretable visualizations using Siamese networks.
problem Creating interpretable visualizations of complex datasets.
method Differentiating Embedding Networks (DEN) using Siamese neural networks and loss functions.
result DEN outperforms existing techniques on FashionMNIST and interpretable features are identified.
t-SNE loses important features in data visualization.
problem t-SNE's loss of important features in data visualization.
method Established mathematical framework to understand t-SNE's loss in different scenarios.
result t-SNE loses important features of data in various scenarios.
Visualizes deep network feature contributions in images.
problem Understanding information flow in deep networks.
method Forward-Backward approach for feature visualization.
result Numerical results show benefits over existing methods.
Survey of methods to visualize neural network features.
problem Understanding neural network activation patterns.
method Activation Maximization and Feature Visualization via Optimization.
result Probabilistic interpretation of AM techniques.
NegToMe uses images to guide text-based models away from unwanted visual elements.
problem Insufficient text-based adversarial guidance for complex visual concepts.
method Negative token merging (NegToMe) using visual features from reference images.
result Significantly enhances output diversity and reduces visual similarity to copyrighted content.
DFF detects similar concepts in images, visualized as heat maps.
problem Localizing similar semantic concepts within images.
method Deep Feature Factorization (DFF) to detect hierarchical cluster structures in feature space.
result Visualizes semantically matching regions across images, revealing network perception.
CNNs reveal retinal ganglion cell features, linking visual processing to neuroscience.
problem Understanding what CNNs learn about retinal neuronal circuits.
method Trained CNNs on white noise images to predict neural responses from salamander retinas.
result CNN filters resemble biological retinal components and ganglion cell receptive fields.
Visual design improves financial data classification accuracy.
problem Improving financial decision-making through better data representation.
method Comparing numeric vs visual data representations in supervised classification.
result Visual transformation of numeric data leads to higher predictability.
We propose a novel methodology, forest floor, to visualize and interpret random forest (RF) models. RF is a popular and useful tool for non-linear multi-variate classification and regression, which yields a good trade-off between robustness (low variance) and adaptiveness (low bias). Direct interpretation of a RF model…
System helps scientists visualize deep learning model of x-ray images.
problem Understanding complex x-ray scattering images with multiple attributes.
method Interactive visualization system in feature space and classification output.
result Users can explore and compare images and attributes flexibly.
Analyzes NFT market trends, trade networks, and visual features.
problem Understanding the structure and evolution of NFT market.
method Data analysis of 6.1 million trades of 4.7 million NFTs.
result NFTs form tight clusters and collections contain visually homogeneous objects.
Improved few-shot visual reasoning with image preprocessing.
problem Few-shot classifiers struggle with abstract visual reasoning tasks.
method Spectral feature removal to emphasize unique image parts.
result Combining spectral preprocessing with Relational Networks improves accuracy nearly 40%.
Visual distance for WordNet synsets using deep learning features.
problem Measuring distances between concepts in WordNet.
method Extract visual features from ImageNet-trained CNNs and use them to represent synsets, defining a new distance measure.
result The proposed visual distance measure outperforms traditional lexical distances.
A new multi-hop FiLM approach improves visual reasoning tasks.
problem Challenging multi-modal tasks like visual question-answering and dialogue.
method Generate FiLM layer parameters in a multi-hop fashion, alternating between attending to language and generating parameters.
result Multi-hop FiLM generation achieves state-of-the-art performance on visual dialogue tasks.
It is becoming increasingly important for machine learning methods to make predictions that are interpretable as well as accurate. In many practical applications, it is of interest which features and feature interactions are relevant to the prediction task. We present a novel method, Selective Bayesian Forest Classifie…
Paper develops a framework for generating coherent image captions using visual features and hierarchical topics.
problem Generating semantically coherent paragraphs to describe image content.
method Plug-and-play hierarchical-topic-guided image paragraph generation framework integrating visual extractor and deep topic model.
result Proposed models can distill interpretable multi-layer semantic topics and generate diverse and coherent captions.
System solves a significant fraction of Bongard problems using visual features and pragmatic reasoning.
problem Solving Bongard problems with intelligent vision systems.
method Image processing, symbolic visual vocabulary, Bayesian inference, pragmatic reasoning.
result Good agreement between induced concepts and Bongard's solutions.
Unified regularization framework for visualizing CNNs.
problem Visualizing concepts learned by convolutional neural networks.
method Mathematical framework unifying regularization methods, Sobolev gradients.
result Sobolev filters provide sharper reconstructions and better control over scales.
JD.com uses a new CNN model to improve ad click prediction.
problem Improving CTR prediction for ads with visual content.
method Proposes Category-specific CNN (CSCNN) to incorporate category knowledge early in the feature extraction process.
result CSCNN outperforms existing methods in CTR prediction.
Combining visual attention with deep reinforcement learning improves sample efficiency.
problem Improving sample efficiency in deep reinforcement learning.
method Visual selective attention mechanism implemented using optical flow and batch normalization.
result Visual selective attention leads to improvements in sample efficiency on Atari games.
Understanding how images of objects and scenes behave in response to specific ego-motions is a crucial aspect of proper visual development, yet existing visual learning methods are conspicuously disconnected from the physical source of their images. We propose to exploit proprioceptive motor signals to provide unsuperv…
A hybrid approach links fMRI data to deep features for visual category decoding.
problem Lack of practical fMRI decoder with CNN structure due to limited brain data.
method Kernel Canonical Correlation Analysis linking fMRI and deep learnt representations.
result Effective in distinguishing semantic visual categories using only brain imaging data.
Enhanced PCA method highlights essential features of clusters in high-dimensional data.
problem Interpreting clusters in dimensionality reduction results is challenging.
method Contrastive Principal Component Analysis (cPCA) for identifying essential features.
result ccPCA method effectively highlights essential features of clusters in high-dimensional data.
New methods for assessing and visualizing feature groups in machine learning models.
problem Lack of methods for interpreting feature groups in machine learning models.
method Permutation-based, refitting, and Shapley-based techniques for grouped feature importance. Introduced a sequential procedure for identifying stable feature combinations. Developed a combined features effect plot.
result Effective methods for assessing and visualizing the importance and effect of feature groups in machine learning models.
The visual systems of many mammals, including humans, is able to integrate the geometric information of visual stimuli and to perform cognitive tasks already at the first stages of the cortical processing. This is thought to be the result of a combination of mechanisms, which include feature extraction at single cell l…
The paper explains knowledge distillation by analyzing visual concepts in DNNs.
problem Understanding how knowledge distillation affects the learning of visual concepts in deep neural networks.
method The paper proposes three hypotheses and designs mathematical metrics to evaluate feature representations of DNNs.
result The hypotheses were verified through experiments on various DNNs.
This work explores how XAI methods can visualize the diversity of feature representations in Bayesian Neural Networks.
problem Explaining the diversity of feature representations learned by Bayesian Neural Networks.
method Application of global XAI methods to visualize and quantify the diversity of feature representations.
result The diversity of learned feature representations correlates with uncertainty estimates and network width.
Improved video and movie description using multitask learning.
problem Lack of training data and poor generalization in video captioning.
method Multitask learning encoder-decoder framework for video sequences.
result Improved performance on multi-caption and single-caption datasets.
A new visualization tool MD plot discovers interesting structures in continuous features.
problem Identifying interesting structures in data distributions, especially with skewed, clipped, or multimodal distributions.
method Proposes a new visualization tool called the mirrored density plot (MD plot) that does not require adjusting density estimation parameters.
result The MD plot outperforms conventional methods in identifying structures in complex distributions.
A new method embeds visual features into semantic space for open-set recognition.
problem Learning unseen classes in open-set recognition.
method Vocabulary-informed Extreme Value Learning (ViEVL) combining EVL and ViL.
result ViEVL embeds visual features into semantic space probabilistically, solving open-set recognition.
Paper proposes redundancy-free features for zero-shot object recognition.
problem Redundant visual features degrade zero-shot object recognition.
method Project original features into a new, statistically independent space.
result RFF-GZSL achieves competitive results on benchmark datasets.
3D-CNN method visualizes localized geometric features for manufacturability analysis.
problem Interpreting 3D-CNN decisions for complex geometries.
method 3D-CNN with surface normals, 3D-GradCAM for feature visualization.
result Identifies critical local features for manufacturability.
A framework visualizes embedding spaces of neural survival analysis models using anchor directions.
problem Visualizing complex embeddings in neural survival analysis models.
method Estimating anchor directions through clustering or user-supplied concepts, revealing relationships with raw inputs and survival times.
result Visualization strategies reveal how anchor directions relate to raw clinical features and survival time distributions.
Simple method disentangles content and style from pre-trained vision models.
problem Learning interpretable features in visual representations.
method Probabilistic linear entanglement model and simple disentanglement algorithm.
result Method provably disentangles content and style features.
Paper explores LSTM visualization techniques for ECGs.
problem Visualizing LSTM models for ECG classification.
method Four visualization techniques applied to ECGs, focusing on input deletion masks.
result Best technique is learning an input deletion mask to reduce class score.
Rotation-equivariant CNN reveals common features in V1 neurons.
problem V1 models fail to predict natural stimuli responses accurately.
method Rotation-equivariant convolutional neural network model.
result Rotation-equivariant network outperforms regular CNN and reveals common features.
Model improves information transfer from visual streams.
problem Challenges in unsupervised learning from continuous visual data.
method Inspired by physics, maximizes mutual information through temporal process.
result Focus of attention enhances information transfer from input stream.
AV-ASR system improves speech recognition with visual context.
problem Improving speech recognition accuracy with visual information.
method Transformer-based architecture with multiresolution and multimodal training.
result Multiresolution training speeds up convergence and improves WER by 18%.
Capsule networks improve on traditional neural networks by using vector activations.
problem Comparing capsule networks to traditional neural networks to validate their benefits.
method Deep visualization analysis, feature encoding across vector components, and instantiation parameter encoding.
result Capsule features encode information differently and provide benefits in computer vision applications.