Search Results for author:
Found 44 papers, 20 papers with code
Targeting Negative Flips in Active Learning using Validation Sets
In this paper, we discuss improving the performance of active learning algorithms both in terms of prediction accuracy and negative flips.
HEX: Hierarchical Emergence Exploitation in Self-Supervised Algorithms
Ideally, SSL algorithms would take advantage of this hierarchical emergence to have an additional regularization term to account for this local dimensional collapse effect.
Benchmarking Human and Automated Prompting in the Segment Anything Model
Additionally, we demonstrate that performance when using automated methods can be improved by up to 68% via a finetuning approach.
CRACKS: Crowdsourcing Resources for Analysis and Categorization of Key Subsurface faults
However, specialized applications that require expert labels lag in data availability.
Ophthalmic Biomarker Detection: Highlights from the IEEE Video and Image Processing Cup 2023 Student Competition
Balancing personalization and generalization is an important challenge to tackle, as the variation within OCT scans of patients between visits can be minimal while the difference in manifestation of the same disease across different patients may be substantial.
Are Objective Explanatory Evaluation metrics Trustworthy? An Adversarial Analysis
Explainable AI (XAI) has revolutionized the field of deep learning by empowering users to have more trust in neural network models.
Intelligent Multi-View Test Time Augmentation
In this study, we introduce an intelligent Test Time Augmentation (TTA) algorithm designed to enhance the robustness and accuracy of image classification models against viewpoint variations.
Explaining Representation Learning with Perceptual Components
In scenarios, where labels are absent, these importance maps provide more intuitive explanations as they are integral to the human visual system.
Taxes Are All You Need: Integration of Taxonomical Hierarchy Relationships into the Contrastive Loss
In reality, the class label is only one level of a \emph{hierarchy of different semantic relationships known as a taxonomy}.
VOICE: Variance of Induced Contrastive Explanations to quantify Uncertainty in Neural Network Interpretability
We show that every image, network, prediction, and explanatory technique has a unique uncertainty.
Effective Data Selection for Seismic Interpretation through Disagreement
By combining this approach with active learning, a well-known machine learning paradigm for data selection, we arrive at a comprehensive and innovative framework for training set selection in seismic interpretation.
Transitional Uncertainty with Layered Intermediate Predictions
We refer to the underlying preservation mechanism as transitional feature preservation.
Counterfactual Gradients-based Quantification of Prediction Trust in Neural Networks
We observe the following: (i) simple methodologies like negative log likelihood and margin classifiers outperform state-of-the-art uncertainty and out-of-distribution detection techniques for misprediction rates, and (ii) the proposed GradTrust is in the Top-2 performing methods on $37$ of the considered $38$ experimental modalities.
Perceptual Quality-based Model Training under Annotator Label Uncertainty
Annotator label uncertainty manifests in variations of labeling quality.
Understanding and Leveraging the Learning Phases of Neural Networks
The learning dynamics of deep neural networks are not well understood.
FOCAL: A Cost-Aware Video Dataset for Active Learning
In this paper, we introduce the FOCAL (Ford-OLIVES Collaboration on Active Learning) dataset which enables the study of the impact of annotation-cost within a video active learning setting.
Clinical Trial Active Learning
There exists two types of clinical trials: retrospective and prospective.
Clinically Labeled Contrastive Learning for OCT Biomarker Classification
We exploit this relationship by using the clinical data as pseudo-labels for our data without biomarker labels in order to choose positive and negative instances for training a backbone network with a supervised contrastive loss.
Exploiting the Distortion-Semantic Interaction in Fisheye Data
In this work, we present a methodology to shape a fisheye-specific representation space that reflects the interaction between distortion and semantic context present in this data modality.
Probing the Purview of Neural Networks via Gradient Analysis
We define purview as the additional capacity necessary to characterize inference samples that differ from the training data.
Gaussian Switch Sampling: A Second Order Approach to Active Learning
To alleviate this issue, we propose a grounded second-order definition of information content and sample importance within the context of active learning.
Stochastic Surprisal: An inferential measurement of Free Energy in Neural Networks
This paper conjectures and validates a framework that allows for action during inference in supervised neural networks.
Forgetful Active Learning with Switch Events: Efficient Sampling for Out-of-Distribution Data
However, existing strategies directly base the data selection on the data representation of the unlabeled data which is random for OOD samples by definition.
On the Ramifications of Human Label Uncertainty
Our evaluation of existing uncertainty estimation algorithms, with the presence of HLU, indicates the limitations of existing uncertainty metrics and algorithms themselves in response to HLU.
Clinical Contrastive Learning for Biomarker Detection
This is accomplished by leveraging the larger amount of clinical data as pseudo-labels for our data without biomarker labels in order to choose positive and negative instances for training a backbone network with a supervised contrastive loss.
OLIVES Dataset: Ophthalmic Labels for Investigating Visual Eye Semantics
The dataset consists of 1268 near-IR fundus images each with at least 49 OCT scans, and 16 biomarkers, along with 4 clinical labels and a disease diagnosis of DR or DME.
Introspective Learning : A Two-Stage Approach for Inference in Neural Networks
Finally, we ground the proposed machine introspection to human introspection for the application of image quality assessment.
DECAL: DEployable Clinical Active Learning
Hence, active learning techniques that are developed for natural images are insufficient for handling medical data.
Volumetric Supervised Contrastive Learning for Seismic Semantic Segmentation
In seismic interpretation, pixel-level labels of various rock structures can be time-consuming and expensive to obtain.
Gradient-Based Adversarial and Out-of-Distribution Detection
We propose to utilize gradients for detecting adversarial and out-of-distribution samples.
Explanatory Paradigms in Neural Networks
With $P$ as the prediction from a neural network, these questions are `Why P?
Contrastive Reasoning in Neural Networks
In this paper, we formalize the structure of contrastive reasoning and propose a methodology to extract a neural network's notion of contrast.
Extracting Causal Visual Features for Limited label Classification
Neural networks trained to classify images do so by identifying features that allow them to distinguish between classes.
Novelty Detection Through Model-Based Characterization of Neural Networks
To articulate the significance of the model perspective in novelty detection, we utilize backpropagated gradients.
Implicit Saliency in Deep Neural Networks
In this paper, we show that existing recognition and localization deep architectures, that have not been exposed to eye tracking data or any saliency datasets, are capable of predicting the human visual saliency.
Contrastive Explanations in Neural Networks
Current modes of visual explanations answer questions of the form $`Why \text{ } P?'$.
Backpropagated Gradient Representations for Anomaly Detection
Anomalies require more drastic model updates to fully represent them compared to normal data.
Characterizing Missing Information in Deep Networks Using Backpropagated Gradients
To complement the learned information from activation-based representation, we propose utilizing a gradient-based representation that explicitly focuses on missing information.
Leveraging Adversarial Examples to Obtain Robust Second-Order Representations
Such a positioning scheme is based on a data point’s second-order property.
Distorted Representation Space Characterization Through Backpropagated Gradients
In this paper, we utilize weight gradients from backpropagation to characterize the representation space learned by deep learning algorithms.
Semantically Interpretable and Controllable Filter Sets
In this paper, we generate and control semantically interpretable filters that are directly learned from natural images in an unsupervised fashion.
Generating Adaptive and Robust Filter Sets Using an Unsupervised Learning Framework
While assessing image quality, the filters need to capture perceptual differences based on dissimilarities between a reference image and its distorted version.
MS-UNIQUE: Multi-model and Sharpness-weighted Unsupervised Image Quality Estimation
We use multiple linear decoders to capture different abstraction levels of the image patches.
CURE-TSR: Challenging Unreal and Real Environments for Traffic Sign Recognition
We benchmark the performance of existing solutions in real-world scenarios and analyze the performance variation with respect to challenging conditions.
