Search Results for author:
Found 30 papers, 13 papers with code
On the Statistical and Information Theoretical Characteristics of DNN Representations
It has been common to argue or imply that a regularizer can be used to alter a statistical property of a hidden layer's representation and thus improve generalization or performance of deep networks.
An Image Grid Can Be Worth a Video: Zero-shot Video Question Answering Using a VLM
Recently, an alternative strategy has surfaced, employing readily available foundation models, such as VideoLMs and LLMs, across multiple stages for modality bridging.
Ranked #1 on
Zero-Shot Video Question Answer
on IntentQA
Improving Forward Compatibility in Class Incremental Learning by Increasing Representation Rank and Feature Richness
Consequently, RFR achieves dual objectives in backward and forward compatibility: minimizing feature extractor modifications and enhancing novel task performance, respectively.
Selectively Informative Description can Reduce Undesired Embedding Entanglements in Text-to-Image Personalization
In text-to-image personalization, a timely and crucial challenge is the tendency of generated images overfitting to the biases present in the reference images.
Harmonizing Visual and Textual Embeddings for Zero-Shot Text-to-Image Customization
In a surge of text-to-image (T2I) models and their customization methods that generate new images of a user-provided subject, current works focus on alleviating the costs incurred by a lengthy per-subject optimization.
On-Off Pattern Encoding and Path-Count Encoding as Deep Neural Network Representations
Specifically, we consider \textit{On-Off pattern} and \textit{PathCount} for investigating how information is stored in deep representations.
Enhancing Contrastive Learning with Efficient Combinatorial Positive Pairing
In the past few years, contrastive learning has played a central role for the success of visual unsupervised representation learning.
A Differentiable Framework for End-to-End Learning of Hybrid Structured Compression
In this study, we develop a \textit{Differentiable Framework~(DF)} that can express filter selection, rank selection, and budget constraint into a single analytical formulation.
Towards a Rigorous Analysis of Mutual Information in Contrastive Learning
Contrastive learning has emerged as a cornerstone in recent achievements of unsupervised representation learning.
VNE: An Effective Method for Improving Deep Representation by Manipulating Eigenvalue Distribution
Since the introduction of deep learning, a wide scope of representation properties, such as decorrelation, whitening, disentanglement, rank, isotropy, and mutual information, have been studied to improve the quality of representation.
Ranked #15 on
Domain Generalization
on TerraIncognita
Meta-Learning with a Geometry-Adaptive Preconditioner
In this study, we propose Geometry-Adaptive Preconditioned gradient descent (GAP) that can overcome the limitations in MAML; GAP can efficiently meta-learn a preconditioner that is dependent on task-specific parameters, and its preconditioner can be shown to be a Riemannian metric.
DR.CPO: Diversified and Realistic 3D Augmentation via Iterative Construction, Random Placement, and HPR Occlusion
To improve the diversity of the whole-body object construction, we develop an iterative method that stochastically combines multiple objects observed from the real world into a single object.
Evaluating Feature Attribution Methods for Electrocardiogram
The performance of cardiac arrhythmia detection with electrocardiograms(ECGs) has been considerably improved since the introduction of deep learning models.
Isotropic Representation Can Improve Dense Retrieval
In particular, many of the high-performing dense retrieval models evaluate representations of query and document using BERT, and subsequently apply a cosine-similarity based scoring to determine the relevance.
Finding Inverse Document Frequency Information in BERT
The traditional approach, however, is being rapidly replaced by Neural Ranking Models (NRMs) that can exploit semantic features.
B2EA: An Evolutionary Algorithm Assisted by Two Bayesian Optimization Modules for Neural Architecture Search
The early pioneering Neural Architecture Search (NAS) works were multi-trial methods applicable to any general search space.
A Highly Effective Low-Rank Compression of Deep Neural Networks with Modified Beam-Search and Modified Stable Rank
The resulting BSR (Beam-search and Stable Rank) algorithm requires only a single hyperparameter to be tuned for the desired compression ratio.
Semi-Siamese Bi-encoder Neural Ranking Model Using Lightweight Fine-Tuning
The results confirm that both lightweight fine-tuning and semi-Siamese are considerably helpful for improving BERT-based bi-encoders.
Mutual Information Estimation as a Difference of Entropies for Unsupervised Representation Learning
In this work, we derive a principled non-contrastive method where mutual information is estimated as a difference of entropies and thus no need for negative sampling.
AID-Purifier: A Light Auxiliary Network for Boosting Adversarial Defense
AID-purifier is an auxiliary network that works as an add-on to an already trained main classifier.
Improving Bi-encoder Document Ranking Models with Two Rankers and Multi-teacher Distillation
When monoBERT is used as the cross-encoder teacher, together with either TwinBERT or ColBERT as the bi-encoder teacher, TRMD produces a student bi-encoder that performs better than the corresponding baseline bi-encoder.
Short-term Traffic Prediction with Deep Neural Networks: A Survey
2) We briefly explain a wide range of DNN techniques from the earliest networks, including Restricted Boltzmann Machines, to the most recent, including graph-based and meta-learning networks.
Interpreting Neural Ranking Models using Grad-CAM
However, explaining the ranking results has become even more difficult with NRM due to the complex structure of neural networks.
DDP-GCN: Multi-Graph Convolutional Network for Spatiotemporal Traffic Forecasting
In this paper, we identify two essential spatial dependencies in traffic forecasting in addition to distance, direction and positional relationship, for designing basic graph elements as the fundamental building blocks.
DEEP-BO for Hyperparameter Optimization of Deep Networks
The performance of deep neural networks (DNN) is very sensitive to the particular choice of hyper-parameters.
Subtask Gated Networks for Non-Intrusive Load Monitoring
In this work, we focus on the idea that appliances have on/off states, and develop a deep network for further performance improvements.
Statistical Characteristics of Deep Representations: An Empirical Investigation
In this study, the effects of eight representation regularization methods are investigated, including two newly developed rank regularizers (RR).
Utilizing Class Information for Deep Network Representation Shaping
Motivated by the idea, we design two class-wise regularizers that explicitly utilize class information: class-wise Covariance Regularizer (cw-CR) and class-wise Variance Regularizer (cw-VR).
Restructuring Batch Normalization to Accelerate CNN Training
Batch Normalization (BN) has become a core design block of modern Convolutional Neural Networks (CNNs).
DNN Representations as Codewords: Manipulating Statistical Properties via Penalty Regularization
Performance of Deep Neural Network (DNN) heavily depends on the characteristics of hidden layer representations.
