In spite of remarkable success of the convolutional neural networks on semantic segmentation, they suffer from catastrophic forgetting: a significant performance drop for the already learned classes when new classes are added on the data, having no annotations for the old classes. We propose an incremental learning met…
On-device research index
arXiv research
A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
168,742 papers · 148 categories
Trend · papers per month
48 results for “new classes”
At present, object recognition studies are mostly conducted in a closed lab setting with classes in test phase typically in training phase. However, real-world problem is far more challenging because: i) new classes unseen in the training phase can appear when predicting; ii) discriminative features need to evolve when…
New boundary constructed for mapping class group.
problem Understanding the structure of mapping class group.
method Action on space of measured foliations to construct new boundary.
result Description of closure of orbit in Thurston and Gardiner-Masur compactifications.
In recent years, more machine learning algorithms have been applied to odor classification. These odor classification algorithms usually assume that the training datasets are static. However, for some odor recognition tasks, new odor classes continually emerge. That is, the odor datasets are dynamically growing while b…
New class of singular complex manifolds studied with degenerate theory.
problem Understanding singular complex manifolds.
method Developed degenerate Kodaira-Hodge theory for new class.
result New degenerate theory for singular complex manifolds.
Proposes a method to estimate acceptance regions for many classes, including new ones.
problem Lack of methods to handle new classes in set-valued classification.
method Generalized Prediction Set (GPS) approach to estimate acceptance regions.
result Achieves a good balance between accuracy, efficiency, and anomaly detection.
System discovers new classes from unlabeled data, improving model performance.
problem Handling datapoints outside initial training distribution.
method Develops new classes through semi-supervised learning, using Dataset Reconstruction Accuracy and class learnability.
result Demonstrates improved model quality through automatic class discovery.
Paper tackles few-shot class-incremental learning with a neural gas network.
problem Incrementally learn new classes from very few labelled samples without forgetting old classes.
method Proposes TOPIC framework using a neural gas network to preserve class topology and adapt to new samples.
result Significantly outperforms other methods on CIFAR100, miniImageNet, and CUB200 datasets.
New minimal link diagrams found, including torus links and homogeneous ones.
problem Finding minimal link diagrams with new classes.
method Morton-Franks-Williams inequality approach.
result New classes of minimal link diagrams, including previously unproven ones.
New examples of subgroups in mapping class groups are found.
problem Understanding subgroups in mapping class groups.
method Constructing new families of parabolically geometrically finite subgroups.
result These subgroups are undistorted in Mod().
New model classes for function approximation by neural networks defined on domains.
problem Defining novel model classes for function approximation on bounded domains.
method Introducing weighted variation spaces to define new model classes on domains.
result New model classes are strictly larger than classical ones but maintain the same NNA rates.
This paper introduces a new perspective on multi-class ensemble classification that considers training an ensemble as a state estimation problem. The new perspective considers the final ensemble classifier model as a static state, which can be estimated using a Kalman filter that combines noisy estimates made by indivi…
New stability theorem for nonorientable surfaces mapping class groups.
problem Stability of homology groups of mapping class groups of nonorientable surfaces.
method Galatius--Kupers--Randal-Williams framework of cellular -algebras.
result New best known stability range for homology of nonorientable surfaces.
A new oversampling framework generates minority samples by perturbing majority classes.
problem Oversampling in imbalanced classification often neglects majority classes, leading to samples spread across the minority space.
method Introduces a counterfactual objective to generate new minority samples by perturbing majority samples.
result Generated minority samples are near the decision boundary and significantly outperform state-of-the-art methods.
We describe a general procedure to produce fundamental domains for complex hyperbolic triangle groups, a class of groups that contains a representative of the commensurability class of every known non-arithmetic lattice in . We discuss several commensurability invariants for lattices, and show that some …
Paper develops a deep neural network for open set incremental learning of new authors.
problem Classifying unseen examples from previously unseen classes.
method Deep neural network clustering and retraining for new classes.
result Incremental learning model that continuously learns new classes.
DRAGON improves learning for rare classes in unbalanced datasets using class descriptions.
problem Learning rare classes in unbalanced datasets with deep models.
method DRAGON is a late-fusion architecture that corrects bias towards frequent classes and fuses class-descriptions to improve tail-class accuracy.
result DRAGON outperforms state-of-the-art models on new benchmarks for long-tail learning with class descriptors.
New Finsler metrics constructed from -metrics.
problem Exploring new Finsler metrics within the -metric class.
method Constructing new sub-classes of -metrics.
result Presented illustrative examples of new Finsler metrics.
The well-known fact that any genus symplectic Lefschetz fibration is given by a word that is equal to the identity element in the mapping class group and each of whose elements is given by a positive Dehn twist, provides an intimate relationship between words in the mapping class group and 4-manif…
Nonnegative Matrix Factorization (NMF) has been a popular representation method for pattern classification problem. It tries to decompose a nonnegative matrix of data samples as the product of a nonnegative basic matrix and a nonnegative coefficient matrix, and the coefficient matrix is used as the new representation. …
Optimal transport method rejects new classes and adjusts class ratios for open set domain adaptation.
problem Handling new classes in target domains with distribution shifts.
method Two-step optimal transport approach: reject new classes first, then adjust class ratios.
result Outperforms state-of-the-art methods in open set domain adaptation.
Enhances model's ability to distinguish target domain by adding a new class.
problem Improving unsupervised domain adaptation models' discriminative power.
method Training model on data from a new class generated by GAN, repositioning current class data.
result Achieves state-of-the-art performance in various unsupervised domain adaptation scenarios.
New conformal prediction methods for long-tailed classification problems.
problem Rare classes are systematically omitted in existing conformal prediction methods.
method Introduced a new conformal score function and a new interpolation procedure.
result Smoothly trade off set size and class-conditional coverage.
Proposes a method to adapt to new classes in a domain shift.
problem Learning new classes in a domain shift without labeled supervision.
method Inspired by prototypical networks, the method classifies target samples into shared and novel classes.
result Superior performance compared to DA and CI methods in the CIDA paradigm.
We present a new method for manufacturing complex-valued harmonic morphisms from a wide class of Riemannian Lie groups. This yields new solutions from an important family of homogeneous Hadamard manifolds. We also give a new method for constructing left-invariant foliations on a large class of Lie groups producing harm…
Introduces a new characteristic class for vector bundles with a connection.
problem Tackles the classification of vector bundles with algebraic connections.
method Defines a new characteristic class using a connection and proves its independence of the choice of connection.
result The class is an invariant of the vector bundle and is stronger than the Chern and Euler classes.
The goal of this work is to study the ideals of the Goldman Lie algebra . To do so, we construct an algebra homomorphism from to a simpler algebraic structure, and focus on finding ideals of this new structure instead. The structure can be regarded as either a -module or a -module gen…
Visual Speech Recognition (VSR) is the process of recognizing or interpreting speech by watching the lip movements of the speaker. Recent machine learning based approaches model VSR as a classification problem; however, the scarcity of training data leads to error-prone systems with very low accuracies in predicting un…
New invariants prove existence of Kahler-Einstein metrics on big classes.
problem Existence of Kahler-Einstein metrics on varieties with klt singularities.
method Introducing new invariants and proving a generalization of the Tian-Odaka-Sano Theorem.
result Proves existence of twisted Kahler-Einstein metrics on big classes.
New bicombings found for mapping class groups and Teichmüller spaces.
problem Finding efficient ways to navigate mapping class groups and Teichmüller spaces.
method Explained bicombings via stable cubical intervals in hierarchically hyperbolic spaces.
result Hierarchical hulls are quasi-isometric to finite CAT(0) cube complexes.
New aesthetic curves in equiaffine geometry include the quadratic and logarithmic spiral.
problem Designing aesthetic shapes in equiaffine geometry.
method Introducing a new symmetry (ESA) to characterize planar curves.
result The new class of curves includes the quadratic curve and logarithmic spiral.
A new generative classification strategy outperforms existing methods in class-incremental learning.
problem Incrementally training deep neural networks to recognize new classes is challenging.
method Proposes learning the joint distribution p(x,y) and performing classification using Bayes' rule, implemented with variational autoencoders and importance sampling.
result Performs very well on continual learning benchmarks, outperforming existing baselines.
Despite the breakthroughs achieved by deep learning models in conventional supervised learning scenarios, their dependence on sufficient labeled training data in each class prevents effective applications of these deep models in situations where labeled training instances for a subset of novel classes are very sparse -…
A new sampling method balances imbalanced data using gamma distribution.
problem Imbalanced class distribution in data causes bias in classification algorithms.
method Intelligent resampling of minority class instances via gamma distribution.
result The proposed method outperforms existing techniques on 12 out of 24 datasets.
We produce new cohomology for non-uniform arithmetic lattices using a technique of Millson--Raghunathan. From this, we obtain new characteristic classes of manifold bundles with fiber a closed -dimensional manifold with indefinite intersection form of signature . These classes are defined on …
New SDP method certifies neural network robustness across all classes efficiently.
problem Certifying robustness of neural networks across multiple classes.
method Quadratic model + SDP relaxation + pruning strategy.
result Significant computational speed-up and scalability to large datasets.
A number of important applied problems in engineering, finance and medicine can be formulated as a problem of anomaly detection. A classical approach to the problem is to describe a normal state using a one-class support vector machine. Then to detect anomalies we quantify a distance from a new observation to the const…
We define a new formal Riemannian metric on a conformal class in the context of the -Yamabe problem. Our construction leads to a new variational characterization and a new parabolic flow approach to this problem. Moreover, this variational framework suggests that solutions to this problem are unique in…
Paper introduces new Finsler metrics preserved under projective transformations.
problem Developing new projective invariant in Finsler geometry.
method Formulated weakly generalized Douglas-Weyl equation to generalize Finsler metrics.
result Introduces new subclasses of Finsler metrics: generalized weakly-Weyl and generalized -metrics.
This paper introduces a novel, generic active learning method for one-class classification. Active learning methods play an important role to reduce the efforts of manual labeling in the field of machine learning. Although many active learning approaches have been proposed during the last years, most of them are restri…
A new metric for uncertainty quantification using class collisions.
problem Fine-grained uncertainty quantification in classification problems.
method Introducing the collision matrix and estimating it from one-hot labeled data.
result The collision matrix uniquely recovers the posterior class probability distribution.
New operators in Khovanov-Rozansky homology exhibit symmetry.
problem Symmetry in Khovanov-Rozansky homology.
method Defined new commuting operators and proved satisfies hard Lefschetz property.
result Symmetry in Khovanov-Rozansky homology is confirmed.
New infinite arrangements found in higher dimensions.
problem Finding new infinite arrangements in higher dimensions.
method Endowing Falk complexes with an injective metric.
result New examples of infinite arrangements in dimension .
New inequalities for austere submanifolds established.
problem Normal scalar curvature inequalities on austere submanifolds.
method Proved sharper DDVV-type inequalities on austere subspaces.
result Achieved equality in normal scalar curvature inequality for a specific austere submanifold.
New method classifies 4-component link homotopy using claspers.
problem Classifying link-homotopy classes of 4-component links.
method Using Habiro's clasper theory to modify Levine's algebraic computations.
result More symmetrical and schematic classification of 4-component link homotopy.
We give new upper bounds on the stable commutator lengths of Dehn twists in mapping class groups and new lower bounds on the stable commutator lengths of Dehn twists in hyperelliptic mapping class groups. In particular, we show that the stable commutator lengths of Dehn twists about a nonseparating and a separating cur…
New subgroups of mapping class groups constructed for infinite-type surfaces.
problem Constructing new subgroups of mapping class groups for infinite-type surfaces.
method Utilization of special homeomorphisms called shift maps and multipush maps.
result Countably (and uncountably in certain cases) many non-conjugate embeddings of subgroups into mapping class groups.
A new data augmentation method selects mixed classes based on class distances for better performance.
problem Improving recognition accuracy in object recognition using deep learning.
method Calculates class distances and selects mixed data from suitable classes dynamically.
result Improves recognition performance on general and long-tailed image recognition datasets.