Search Results for author:
Found 25 papers, 9 papers with code
Token Fusion: Bridging the Gap between Token Pruning and Token Merging
In this paper, we introduce "Token Fusion" (ToFu), a method that amalgamates the benefits of both token pruning and token merging.
Continual Diffusion with STAMINA: STack-And-Mask INcremental Adapters
We show that STAMINA outperforms the prior SOTA for the setting of text-to-image continual customization on a 50-concept benchmark composed of landmarks and human faces, with no stored replay data.
Continual Diffusion: Continual Customization of Text-to-Image Diffusion with C-LoRA
We show that C-LoRA not only outperforms several baselines for our proposed setting of text-to-image continual customization, which we refer to as Continual Diffusion, but that we achieve a new state-of-the-art in the well-established rehearsal-free continual learning setting for image classification.
Numerical Optimizations for Weighted Low-rank Estimation on Language Model
However, standard SVD treats the parameters within the matrix with equal importance, which is a simple but unrealistic assumption.
Language model compression with weighted low-rank factorization
In other words, the optimization objective of SVD is not aligned with the trained model's task accuracy.
A Closer Look at Rehearsal-Free Continual Learning
Next, we explore how to leverage knowledge from a pre-trained model in rehearsal-free continual learning and find that vanilla L2 parameter regularization outperforms EWC parameter regularization and feature distillation.
A Closer Look at Knowledge Distillation with Features, Logits, and Gradients
Knowledge distillation (KD) is a substantial strategy for transferring learned knowledge from one neural network model to another.
Hyperparameter-free Continuous Learning for Domain Classification in Natural Language Understanding
Most existing continual learning approaches suffer from low accuracy and performance fluctuation, especially when the distributions of old and new data are significantly different.
Lite-MDETR: A Lightweight Multi-Modal Detector
The key primitive is that Dictionary-Lookup-Transformormations (DLT) is proposed to replace Linear Transformation (LT) in multi-modal detectors where each weight in Linear Transformation (LT) is approximately factorized into a smaller dictionary, index, and coefficient.
A Geometric Perspective towards Neural Calibration via Sensitivity Decomposition
It is well known that vision classification models suffer from poor calibration in the face of data distribution shifts.
Exploring Covariate and Concept Shift for Detection and Confidence Calibration of Out-of-Distribution Data
To this end, we theoretically derive two score functions for OOD detection, the covariate shift score and concept shift score, based on the decomposition of KL-divergence for both scores, and propose a geometrically-inspired method (Geometric ODIN) to improve OOD detection under both shifts with only in-distribution data.
DictFormer: Tiny Transformer with Shared Dictionary
DictFormer significantly reduces the redundancy in the transformer's parameters by replacing the prior transformer's parameters with compact, shared dictionary, a few unshared coefficients, and indices.
Enhancing the Generalization for Intent Classification and Out-of-Domain Detection in SLU
This paper proposes to train a model with only IND data while supporting both IND intent classification and OOD detection.
Always Be Dreaming: A New Approach for Data-Free Class-Incremental Learning
Modern computer vision applications suffer from catastrophic forgetting when incrementally learning new concepts over time.
Ranked #5 on
Class Incremental Learning
on cifar100
Memory-Efficient Semi-Supervised Continual Learning: The World is its Own Replay Buffer
Our work investigates whether we can significantly reduce this memory budget by leveraging unlabeled data from an agent's environment in a realistic and challenging continual learning paradigm.
Posterior Re-calibration for Imbalanced Datasets
Neural Networks can perform poorly when the training label distribution is heavily imbalanced, as well as when the testing data differs from the training distribution.
Ranked #31 on
Long-tail Learning
on CIFAR-100-LT (ρ=10)
Generalized ODIN: Detecting Out-of-distribution Image without Learning from Out-of-distribution Data
Deep neural networks have attained remarkable performance when applied to data that comes from the same distribution as that of the training set, but can significantly degrade otherwise.
Out-of-Distribution Detection
Out of Distribution (OOD) Detection
Multi-class Classification without Multi-class Labels
This work presents a new strategy for multi-class classification that requires no class-specific labels, but instead leverages pairwise similarity between examples, which is a weaker form of annotation.
Re-evaluating Continual Learning Scenarios: A Categorization and Case for Strong Baselines
Continual learning has received a great deal of attention recently with several approaches being proposed.
A probabilistic constrained clustering for transfer learning and image category discovery
The proposed objective directly minimizes the negative log-likelihood of cluster assignment with respect to the pairwise constraints, has no hyper-parameters, and demonstrates improved scalability and performance on both supervised learning and unsupervised transfer learning.
Ranked #1 on
Ecg Risk Stratification
on ngm
Learning to Cluster for Proposal-Free Instance Segmentation
We utilize the most fundamental property of instance labeling -- the pairwise relationship between pixels -- as the supervision to formulate the learning objective, then apply it to train a fully convolutional network (FCN) for learning to perform pixel-wise clustering.
Ranked #13 on
Lane Detection
on TuSimple
Learning to cluster in order to transfer across domains and tasks
The key insight is that, in addition to features, we can transfer similarity information and this is sufficient to learn a similarity function and clustering network to perform both domain adaptation and cross-task transfer learning.
Training Shallow and Thin Networks for Acceleration via Knowledge Distillation with Conditional Adversarial Networks
There is an increasing interest on accelerating neural networks for real-time applications.
Deep Image Category Discovery using a Transferred Similarity Function
We propose that this network can be learned with contrastive loss which is only based on weak binary pair-wise constraints.
Neural network-based clustering using pairwise constraints
Robustness analysis also shows that the method is largely insensitive to the number of clusters.
