Quantum hybrid vision transformers improve event classification in high energy physics.
problem Excessive computational resources for training and deploying vision transformer models.
method Constructed quantum hybrid vision transformers for high energy physics event classification.
result Quantum hybrid models achieve comparable performance to classical models with fewer parameters.
B-cos transformers explain Vision Transformers' decisions.
problem Lack of holistic explanations for transformer outputs.
method Formulate each component as dynamic linear, allowing a single linear transform for summarization.
result Bcos-ViTs are highly interpretable and competitive on ImageNet.
This work characterizes benign overfitting in Vision Transformers.
problem Understanding generalization of Vision Transformers when trained to overfit.
method Gradient descent on a data distribution model, focusing on self-attention layer and softmax.
result Established a condition to distinguish between small and large test errors based on signal-to-noise ratio.
Vision transformers benefit from non-smooth components in adaptation.
problem Understanding the role of non-smoothness in vision transformer adaptation.
method Theoretical analysis and extensive experiments on large-scale vision transformers.
result High plasticity of attention modules and feedforward layers leads to better finetuning performance.
CViT learns complex physical systems using vision transformer techniques.
problem Learning maps between infinite-dimensional function spaces in scientific machine learning.
method Combines vision transformer encoder, grid-based coordinate embedding, and cross-attention mechanism.
result Achieves state-of-the-art performance on multiple benchmarks, often surpassing larger models.
ADAPT improves robustness of Vision Transformers without full model fine-tuning.
problem Vulnerability of Vision Transformers to adversarial attacks.
method Parameter-efficient prompt tuning with ADAPT framework for adaptive adversarial training.
result ADAPT achieves robust accuracy of ~40% w.r.t. SOTA methods using only ~1% of the parameters.
Effective theory for Transformer initialization improves model performance.
problem Improving performance of Transformers at initialization.
method Effective-theory analysis of signal propagation in wide and deep Transformers.
result Particular width scalings of initialization and training hyperparameters.
Vision Transformers show different internal representations compared to CNNs.
problem Understanding how Vision Transformers solve image classification tasks.
method Comparative analysis of ViT and CNN architectures on image classification benchmarks.
result ViT has more uniform representations across all layers, while CNNs have more varied representations.
CNNs encode data augmentation transformations, especially in early layers.
problem Whether neural network features encode data augmentation transformations.
method Systematic approach using pre-trained vision models to predict augmentation transformations.
result Neural network features encode data augmentation transformations, especially in early layers.
CutMix training technique improves spatial locality in Vision Transformers.
problem Improving spatial locality in Vision Transformers trained from scratch.
method Comparison of Baseline and Modern training protocols on CIFAR-10, CIFAR-100, and Tiny-ImageNet.
result CutMix training component significantly reduces Mean Attention Distance (MAD) in early layers of Vision Transformers.
Sparse Vision MoE matches dense networks in image recognition while using less compute.
problem Scaling vision models efficiently in computer vision.
method Vision MoE (V-MoE) - a sparse version of Vision Transformer.
result V-MoE matches state-of-the-art dense networks in image recognition with half the compute.
Mobile V-MoEs scale down ViTs for resource-constrained vision tasks.
problem Scaling down Vision Transformers for resource-constrained applications.
method Sparse Mixture-of-Experts (MoEs) applied to entire images, with a stable training procedure.
result Mobile V-MoEs achieve better performance-efficiency trade-offs than dense ViTs.
Study proposes a statistical test for Vision Transformer's attention mechanisms.
problem ViT's attention mechanisms may focus on irrelevant regions, leading to unreliable evidence.
method Selective inference framework to quantify statistical significance of attentions as p-values.
result Proposed method enables reliable quantification of false positive detection probability of attentions.
ViLT is a faster vision-and-language model without convolution or region supervision.
problem Efficiency and expressive power limitations in current VLP models.
method A convolution-free Vision-and-Language Transformer (ViLT) that processes visual inputs similarly to textual inputs.
result ViLT is up to tens of times faster with competitive or better performance.
GTA improves transformer-based NVS models by encoding geometric structure.
problem Suboptimal positional encoding for 3D vision tasks.
method Geometry-aware attention mechanism encoding geometric structure of tokens.
result GTA improves learning efficiency and performance of NVS models.
Mimetic initialization improves Transformer training on small datasets.
problem Difficulty in training Transformers on small datasets.
method Initialize self-attention layers to look like pre-trained models.
result Vanilla Transformers trained with mimetic initialization achieve higher accuracy.
Novel regularization for Vision Transformers improves model generalization and sparsity.
problem Improving generalization and sparsity in Vision Transformers.
method Likelihood-guided variational Ising-based regularization.
result Improved generalization and sparsity in Vision Transformers.
Theoretical analysis of vision transformers' performance with MAE and CL objectives.
problem Understanding the distinct representations learned by vision transformers with MAE and CL objectives.
method Modeling visual data distribution and analyzing ViTs training dynamics with gradient descent.
result ViTs trained with MAE objectives learn both global and local features, while CL-trained ViTs favor global features.
Develops methods for training models with few annotations in computer vision tasks.
problem Training models with limited labeled data in computer vision.
method Theoretical, algorithmic, and experimental contributions for Meta-Learning and Semi-Supervised Learning.
result Improves Contrastive Learning and proposes a semi-supervised method for transformer-based object detectors.
Evaluates deep learning models in histopathology for robustness and classification strategies.
problem Lack of comprehensive evaluation of histopathology models beyond accuracy.
method Developed a new methodology to evaluate models on five histopathology datasets, including vision transformers and CNNs.
result Identified insights into cancer classification strategies and robustness against stain variations.
Paper proposes privacy-preserving learning for images, making them imperceptible to humans but recognizable by machines.
problem Conflict between developing AI systems and protecting sensitive training data.
method Encryption strategies (random shuffling and sub-patch mixing) followed by minimal adaptation to vision transformer.
result Achieves comparable accuracy to competitive methods while ensuring human-imperceptibility of encrypted images.
Measures equivariance in vision models using Lie derivative.
problem Understanding the role of equivariance in recent vision models.
method Introducing Lie derivative to measure equivariance with strong mathematical foundations and minimal hyperparameters.
result Many violations of equivariance can be linked to spatial aliasing in network layers, and larger models tend to display more equivariance.
This paper analyzes shallow ViTs, providing sample complexity and SGD behavior insights.
problem Theoretical understanding of shallow ViTs, especially their sample complexity and SGD behavior.
method Data model with label-relevant and label-irrelevant tokens, theoretical analysis of shallow ViT training.
result Characterization of sample complexity for zero generalization error in shallow ViTs.
Task-agnostic data augmentation shows little benefit for pretrained transformers.
problem Evaluating the effectiveness of task-agnostic data augmentation on pretrained transformers.
method Conducted a systematic examination of two data augmentation techniques (Easy Data Augmentation and Back-Translation) across 5 tasks, 6 datasets, and 3 pretrained transformer models.
result Data augmentation techniques previously effective for non-pretrained models fail to consistently improve performance for pretrained transformers, even with limited training data.
New method reduces training cost by using approximate gradients.
problem Training neural networks is computationally expensive.
method Uses control variates to approximate gradients without full backward pass.
result Efficacy demonstrated on a vision transformer classification task.
CST-YOLO improves blood cell detection with YOLOv7 and CNN-Swin Transformer.
problem Small-scale object detection in blood cells.
method YOLOv7 architecture enhanced with CNN-Swin Transformer, W-ELAN, MCS, CatConv.
result CST-YOLO achieves 92.7%, 95.6%, and 91.1% mAP@0.5 on three blood cell datasets.
Transformers improve time series modeling by capturing long-range dependencies.
problem Capturing long-range dependencies in time series data.
method Summarized and reviewed adaptations of Transformers for time series analysis.
result Transformers enhance time series forecasting, anomaly detection, and classification.
Paper compresses ViTs models by 60% with minimal accuracy loss.
problem Memory constraints in ViTs models.
method Activation-aware low-rank tensor approximations.
result 60% reduction in model size with <1% accuracy loss.
Regularizes attention scores in vision transformers using bootstrapping.
problem Noisy and diffused attention maps in ViT limit interpretability.
method Statistical learning techniques, bootstrapping of attention scores.
result Improves shrinkage and sparsity of attention scores.
This research integrates attention into XAI frameworks for better model explanations.
problem Improving the interpretability of transformer models.
method Developed two novel explanation methods: Shapley value decomposition and token-level directional derivatives.
result Attention weights can be meaningfully incorporated into XAI frameworks, enhancing transformer explainability.
Fusion of transformer networks using optimal transport for improved performance.
problem Improving performance of transformer-based models through fusion.
method Exploiting optimal transport for soft alignment of transformer components.
result Consistently outperforms vanilla fusion and individual parent models.
Unified model trained on images and videos using masked autoencoding.
problem Training a single model for multiple visual modalities.
method Masked autoencoding on a Vision Transformer.
result Unified model achieves comparable or better performance than single-modality models.
ConViT combines CNN and ViT strengths, improving image classification.
problem Combining the strengths of CNNs and ViTs while avoiding their limitations.
method Introducing GPSA, a form of positional self-attention with a soft convolutional inductive bias.
result ConViT outperforms DeiT on ImageNet while offering improved sample efficiency.
STRING improves 2D and 3D position encodings for better performance.
problem Efficient and accurate position encoding for 2D and 3D applications.
method STRING extends Rotary Position Encodings with a unifying theoretical framework, maintaining translation invariance and low computational cost.
result STRING shows substantial gains in open-vocabulary object detection and robotics.
Transformers interpreted as probabilistic Laplacian Eigenmaps steps.
problem Improving transformer performance through probabilistic interpretation.
method Probabilistic Laplacian Eigenmaps model derivation and graph diffusion step.
result Subtracting identity from attention matrix improves transformer performance.
PACE explains ViTs by modeling patch-level concept distributions, surpassing existing methods.
problem Lack of trustworthy post-hoc explanations for Vision Transformers (ViTs)
method Variational Bayesian explanation framework (PACE)
result PACE surpasses state-of-the-art methods in meeting desiderata for ViT explanations.
Centroid Transformers reduce memory and computation by summarizing inputs into centroids.
problem Efficiently summarize inputs with reduced memory and computation.
method Generalizes self-attention to map N inputs to M centroids (M ≤ N), reducing complexity.
result Centroid Transformers reduce memory and computation while preserving key information.
Few-Shot Diffusion Models generate new samples from small image sets.
problem Generating new samples from a few images.
method Conditional Denoising Diffusion Probabilistic Models (DDPM) with patch-based input set information.
result FSDM can generate samples from previously unseen classes conditioned on as few as 5 samples.
New method improves reliability of depth estimation models.
problem Uncertainty quantification in large-scale vision models.
method Parameter-efficient Bayesian neural networks with PEFT methods.
result Combining PEFT methods with Bayesian inference enhances predictive performance.
The Transformer architecture has become increasingly popular over the past two years, owing to its impressive performance on a number of natural language processing (NLP) tasks. However, all Transformer computations occur at the level of word representations and therefore, it may be argued that Transformer models do no…
Study benchmarks uncertainty quantification in chest X-ray classification.
problem Reliable uncertainty quantification for medical AI models.
method Evaluation of 13 uncertainty quantification methods on MIMIC-CXR-JPG dataset.
result Insights into effectiveness and disentanglement of epistemic and aleatoric uncertainties.
Unified inference framework for spatiotemporal data.
problem Challenges in extracting mechanistic insights from complex spatiotemporal data.
method Vision transformer-driven variational encoding and likelihood-free Bayesian approach.
result Unified inference framework for identifying spatial and temporal patterns.
Paper analyzes why deeper layers of ViTs perform worse on out-of-distribution tasks.
problem Performance degradation of intermediate layers in ViTs under distribution shift.
method Extensive linear probing experiments across various benchmarks and fine-grained analysis of transformer modules.
result Probing feedforward network activations yields best performance under significant distribution shift.
LEMON uses pre-trained models to scale neural networks efficiently.
problem Efficiency in scaling deep neural networks, especially Transformers, which are resource-intensive to train from scratch.
method LEMON initializes scaled models using pre-trained weights and optimizes learning rates.
result Significant reduction in training time and computational costs for Vision Transformers and BERT.
Transformers can predict new tokens based on any number of context tokens, approximating continuous mappings with fixed resources.
problem Handling an arbitrarily large number of context tokens in transformers.
method Mathematical analysis of transformer's expressivity using Wasserstein distance and continuous mappings.
result Deep transformers are universal and can approximate continuous in-context mappings to arbitrary precision, uniformly over compact token domains.
CDAM improves attention maps for ViTs, making them more class-sensitive.
problem Existing attention maps in ViTs lack class sensitivity.
method Class-discriminative attention maps (CDAM) that scale attention scores by class relevance.
result CDAM provides more class-sensitive explanations than existing methods.
SGPA calibrates transformer uncertainty for safety-critical tasks.
problem Uncertainty estimation in transformer models for safety-critical domains.
method Bayesian inference in transformer's output space using sparse Gaussian processes.
result SGPA-based Transformers improve in-distribution calibration and out-of-distribution robustness.
The alignment of a set of objects by means of transformations plays an important role in computer vision. Whilst the case for only two objects can be solved globally, when multiple objects are considered usually iterative methods are used. In practice the iterative methods perform well if the relative transformations b…