Search Results for author:
Found 31 papers, 12 papers with code
Multi-domain improves out-of-distribution and data-limited scenarios for medical image analysis
We refer to this approach as multi-domain model and compare its performance to that of specialized models.
D3: Data Diversity Design for Systematic Generalization in Visual Question Answering
We demonstrate that this result is independent of the similarity between the training and testing data and applies to well-known families of neural network architectures for VQA (i. e. monolithic architectures and neural module networks).
Modularity Trumps Invariance for Compositional Robustness
In this work we develop a compositional image classification task where, given a few elemental corruptions, models are asked to generalize to compositions of these corruptions.
Deephys: Deep Electrophysiology, Debugging Neural Networks under Distribution Shifts
Deep Neural Networks (DNNs) often fail in out-of-distribution scenarios.
Transformer Module Networks for Systematic Generalization in Visual Question Answering
We reveal that Neural Module Networks (NMNs), i. e., question-specific compositions of modules that tackle a sub-task, achieve better or similar systematic generalization performance than the conventional Transformers, even though NMNs' modules are CNN-based.
Do Neural Networks for Segmentation Understand Insideness?
Only recurrent networks trained with small images learn solutions that generalize well to almost any curve.
Robust Upper Bounds for Adversarial Training
We introduce a new approach to adversarial training by minimizing an upper bound of the adversarial loss that is based on a holistic expansion of the network instead of separate bounds for each layer.
Symmetry Perception by Deep Networks: Inadequacy of Feed-Forward Architectures and Improvements with Recurrent Connections
Symmetry is omnipresent in nature and perceived by the visual system of many species, as it facilitates detecting ecologically important classes of objects in our environment.
Three approaches to facilitate DNN generalization to objects in out-of-distribution orientations and illuminations
While humans have a remarkable capability of recognizing objects in out-of-distribution (OoD) orientations and illuminations, Deep Neural Networks (DNNs) severely suffer in this case, even when large amounts of training examples are available.
Emergent Neural Network Mechanisms for Generalization to Objects in Novel Orientations
The capability of Deep Neural Networks (DNNs) to recognize objects in orientations outside the distribution of the training data is not well understood.
The Foes of Neural Network's Data Efficiency Among Unnecessary Input Dimensions
Datasets often contain input dimensions that are unnecessary to predict the output label, e. g. background in object recognition, which lead to more trainable parameters.
Adversarial examples within the training distribution: A widespread challenge
This result provides evidence supporting theories attributing adversarial examples to the proximity of data to ground-truth class boundaries, and calls into question other theories which do not account for this more stringent definition of adversarial attacks.
How Modular Should Neural Module Networks Be for Systematic Generalization?
Neural Module Networks (NMNs) aim at Visual Question Answering (VQA) via composition of modules that tackle a sub-task.
On the Capability of CNNs to Generalize to Unseen Category-Viewpoint Combinations
We find that learning category and viewpoint in separate networks compared to a shared one leads to an increase in selectivity and invariance, as separate networks are not forced to preserve information about both category and viewpoint.
The Foes of Neural Network’s Data Efficiency Among Unnecessary Input Dimensions
In this paper, we investigate the impact of unnecessary input dimensions on one of the central issues of machine learning: the number of training examples needed to achieve high generalization performance, which we refer to as the network's data efficiency.
When and how CNNs generalize to out-of-distribution category-viewpoint combinations
In this paper, we investigate when and how such OOD generalization may be possible by evaluating CNNs trained to classify both object category and 3D viewpoint on OOD combinations, and identifying the neural mechanisms that facilitate such OOD generalization.
Robustness to Transformations Across Categories: Is Robustness To Transformations Driven by Invariant Neural Representations?
Our results with state-of-the-art DCNNs indicate that invariant neural representations do not always drive robustness to transformations, as networks show robustness for categories seen transformed during training even in the absence of invariant neural representations.
Frivolous Units: Wider Networks Are Not Really That Wide
We identify two distinct types of "frivolous" units that proliferate when the network's width is increased: prunable units which can be dropped out of the network without significant change to the output and redundant units whose activities can be expressed as a linear combination of others.
Do Deep Neural Networks for Segmentation Understand Insideness?
At the heart of image segmentation lies the problem of determining whether a pixel is inside or outside a region, which we denote as the "insideness" problem.
Minimal Images in Deep Neural Networks: Fragile Object Recognition in Natural Images
Ullman et al. 2016 show that a slight modification of the location and size of the visible region of the minimal image produces a sharp drop in human recognition accuracy.
Theory IIIb: Generalization in Deep Networks
Here we prove a similar result for nonlinear multilayer DNNs near zero minima of the empirical loss.
Theory of Deep Learning III: explaining the non-overfitting puzzle
In this note, we show that the dynamics associated to gradient descent minimization of nonlinear networks is topologically equivalent, near the asymptotically stable minima of the empirical error, to linear gradient system in a quadratic potential with a degenerate (for square loss) or almost degenerate (for logistic or crossentropy loss) Hessian.
Herding Generalizes Diverse M -Best Solutions
We show that the algorithm to extract diverse M -solutions from a Conditional Random Field (called divMbest [1]) takes exactly the form of a Herding procedure [2], i. e. a deterministic dynamical system that produces a sequence of hypotheses that respect a set of observed moment constraints.
Predicting When Saliency Maps Are Accurate and Eye Fixations Consistent
Many computational models of visual attention use image features and machine learning techniques to predict eye fixation locations as saliency maps.
SALICON: Reducing the Semantic Gap in Saliency Prediction by Adapting Deep Neural Networks
Saliency in Context (SALICON) is an ongoing effort that aims at understanding and predicting visual attention.
Foveation-based Mechanisms Alleviate Adversarial Examples
To see this, first, we report results in ImageNet that lead to a revision of the hypothesis that adversarial perturbations are a consequence of CNNs acting as a linear classifier: CNNs act locally linearly to changes in the image regions with objects recognized by the CNN, and in other regions the CNN may act non-linearly.
Self-Adaptable Templates for Feature Coding
Within this processing pipeline, the common trend is to learn the feature coding templates, often referred as codebook entries, filters, or over-complete basis.
Comment on "Ensemble Projection for Semi-supervised Image Classification"
In a series of papers by Dai and colleagues [1, 2], a feature map (or kernel) was introduced for semi- and unsupervised learning.
SEEDS: Superpixels Extracted via Energy-Driven Sampling
We define a robust and fast to evaluate energy function, based on enforcing color similarity between the bound- aries and the superpixel color histogram.
Random Binary Mappings for Kernel Learning and Efficient SVM
SVMs suffer from various drawbacks in terms of selecting the right kernel, which depends on the image descriptors, as well as computational and memory efficiency.
Sparse Quantization for Patch Description
We demonstrate the capabilities of our formulation for both keypoint matching and image classification.
