arXiv research

Simple 1D-CNN network predicts electricity loads 36 hours ahead.

New approximative kernels improve PDE-G-CNNs for geometric deep learning.

Paper learns an explainer to interpret CNN features without annotations.

In recent years, deep learning poses a deep technical revolution in almost every field and attracts great attentions from industry and academia. Especially, the convolutional neural network (CNN), one representative model of deep learning, achieves great successes in computer vision and natural language processing. How…

Large receptive field CNNs improve distant speech recognition.

L-CNNs preserve gauge symmetry in neural networks.

Unified framework explains geometric properties of CNNs.

Numerous pattern recognition applications can be formed as learning from graph-structured data, including social network, protein-interaction network, the world wide web data, knowledge graph, etc. While convolutional neural network (CNN) facilitates great advances in gridded image/video understanding tasks, very limit…

CNN-F uses generative feedback to improve neural networks' robustness to perturbations.

AT-CNNs show improved shape recognition over texture recognition.

L-CNNs maintain gauge symmetry on non-Abelian lattice theories.

LeanResNet reduces CNN complexity without sacrificing accuracy.

Paper analyzes the free energy of CNNs with skip connections in Bayesian learning.

Theoretical analysis of CNNs' inductive biases and their efficiency in approximating functions.

The paper analyzes deep ReLU CNNs' approximation properties in 2D space.

Combines CNN and LSTM for spatio-temporal graph networks.

New bounds for CNNs show better generalization than previous models.

New multigrid approach reduces CNN parameters by focusing on structured convolutions.

Proposes a curriculum-based scheme to smooth CNN feature embeddings.

A hybrid CNN with shared parameters learns to incorporate recurrent structures.

L-CNNs preserve gauge symmetry in lattice simulations.

2D CNNs approximate Korobov functions with near-optimal rates.

Study on CNNs' learning rates and approximation capacities.

Inverse problems in imaging such as denoising, deblurring, superresolution (SR) have been addressed for many decades. In recent years, convolutional neural networks (CNNs) have been widely used for many inverse problem areas. Although their indisputable success, CNNs are not mathematically validated as to how and what …

LeanConvNets reduces CNN costs without sacrificing accuracy.

SCNN improves video object detection speed by 178%.

It is widely believed that the practical success of Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs) owes to the fact that CNNs and RNNs use a more compact parametric representation than their Fully-Connected Neural Network (FNN) counterparts, and consequently require fewer training examples to…

This paper introduces a way to learn cross-modal convolutional neural network (X-CNN) architectures from a base convolutional network (CNN) and the training data to reduce the design cost and enable applying cross-modal networks in sparse data environments. Two approaches for building X-CNNs are presented. The base app…

GWNN uses graph wavelets for efficient graph CNNs.

Deep CNN classifies EEG-based brain connectivity in schizophrenia.

Improves CNN stability by translating classical signal denoising methods.

Paper studies the expressivity of Convolutional Neural Networks (CNNs).

In this paper, we propose a deep multimodal fusion network to fuse multiple modalities (face, iris, and fingerprint) for person identification. The proposed deep multimodal fusion algorithm consists of multiple streams of modality-specific Convolutional Neural Networks (CNNs), which are jointly optimized at multiple fe…

Group equivariant and steerable convolutional neural networks (regular and steerable G-CNNs) have recently emerged as a very effective model class for learning from signal data such as 2D and 3D images, video, and other data where symmetries are present. In geometrical terms, regular G-CNNs represent data in terms of s…

This paper reports the performances of shallow word-level convolutional neural networks (CNN), our earlier work (2015), on the eight datasets with relatively large training data that were used for testing the very deep character-level CNN in Conneau et al. (2016). Our findings are as follows. The shallow word-level CNN…

Paper proposes CNN-LSTM model for multi-speaker speech separation.

This paper proposes a principled information theoretic analysis of classification for deep neural network structures, e.g. convolutional neural networks (CNN). The output of convolutional filters is modeled as a random variable Y conditioned on the object class C and network filter bank F. The conditional entropy (CENT…

PDE-based G-CNNs add geometric symmetries to CNNs without augmentation.

The paper improves DFA for CNN and RNN training to match BP accuracy.

Whereas CNNs have demonstrated immense progress in many vision problems, they suffer from a dependence on monumental amounts of labeled training data. On the other hand, dictionary learning does not scale to the size of problems that CNNs can handle, despite being very effective at low-level vision tasks such as denois…

Improved robustness of 1D CNNs for heart arrhythmia classification.

HOTCAKE compresses CNNs by decomposing kernels into smaller parts.

Bayesian regularizations are explicitly implemented in CNNs, improving deep learning generalization.

In this paper we propose cross-modal convolutional neural networks (X-CNNs), a novel biologically inspired type of CNN architectures, treating gradient descent-specialised CNNs as individual units of processing in a larger-scale network topology, while allowing for unconstrained information flow and/or weight sharing b…

New tensor network decompositions improve CNN performance.

This paper proposes a framework based on deep convolutional neural networks (CNNs) for automatic heart sound classification using short-segments of individual heart beats. We design a 1D-CNN that directly learns features from raw heart-sound signals, and a 2D-CNN that takes inputs of two- dimensional time-frequency fea…

The paper introduces a method to explain redundancy in deep CNNs using unit impulse response.

ResNet-type CNNs achieve optimal error rates in function classes with block-sparse structures.