Search Results for author:
Found 126 papers, 86 papers with code
Where and when to look? Spatial-temporal attention for action recognition in videos
Our model is efficient, as it proposes a separable spatio-temporal mechanism for video attention, while being able to identify important parts of the video both spatially and temporally.
Action Recognition In Videos
Temporal Action Localization
+1
Exploring Text-to-Motion Generation with Human Preference
This paper presents an exploration of preference learning in text-to-motion generation.
LLaVA-UHD: an LMM Perceiving Any Aspect Ratio and High-Resolution Images
To address the challenges, we present LLaVA-UHD, a large multimodal model that can efficiently perceive images in any aspect ratio and high resolution.
Dynamic Tuning Towards Parameter and Inference Efficiency for ViT Adaptation
Existing parameter-efficient fine-tuning (PEFT) methods have achieved significant success on vision transformers (ViTs) adaptation by improving parameter efficiency.
GRA: Detecting Oriented Objects through Group-wise Rotating and Attention
GRA can adaptively capture fine-grained features of objects with diverse orientations, comprising two key components: Group-wise Rotating and Group-wise Attention.
Glyph-ByT5: A Customized Text Encoder for Accurate Visual Text Rendering
Visual text rendering poses a fundamental challenge for contemporary text-to-image generation models, with the core problem lying in text encoder deficiencies.
2023 Low-Power Computer Vision Challenge (LPCVC) Summary
This article describes the 2023 IEEE Low-Power Computer Vision Challenge (LPCVC).
Probabilistic Contrastive Learning for Long-Tailed Visual Recognition
To overcome this obstacle, we propose a novel probabilistic contrastive (ProCo) learning algorithm that estimates the data distribution of the samples from each class in the feature space, and samples contrastive pairs accordingly.
Ranked #8 on
Long-tail Learning
on iNaturalist 2018
SimPro: A Simple Probabilistic Framework Towards Realistic Long-Tailed Semi-Supervised Learning
Recent advancements in semi-supervised learning have focused on a more realistic yet challenging task: addressing imbalances in labeled data while the class distribution of unlabeled data remains both unknown and potentially mismatched.
LLM Agents for Psychology: A Study on Gamified Assessments
Psychological measurement is essential for mental health, self-understanding, and personal development.
Segment3D: Learning Fine-Grained Class-Agnostic 3D Segmentation without Manual Labels
Therefore, we explore the use of image segmentation foundation models to automatically generate training labels for 3D segmentation.
A-SDM: Accelerating Stable Diffusion through Redundancy Removal and Performance Optimization
The Stable Diffusion Model (SDM) is a popular and efficient text-to-image (t2i) generation and image-to-image (i2i) generation model.
Mask Grounding for Referring Image Segmentation
To tackle this challenge, we introduce a novel Mask Grounding auxiliary task that significantly improves visual grounding within language features, by explicitly teaching the model to learn fine-grained correspondence between masked textual tokens and their matching visual objects.
Ranked #2 on
Referring Expression Segmentation
on RefCOCO testB
GSVA: Generalized Segmentation via Multimodal Large Language Models
Generalized Referring Expression Segmentation (GRES) extends the scope of classic RES to refer to multiple objects in one expression or identify the empty targets absent in the image.
Generalized Referring Expression Segmentation
Referring Expression
+1
Agent Attention: On the Integration of Softmax and Linear Attention
Specifically, the Agent Attention, denoted as a quadruple $(Q, A, K, V)$, introduces an additional set of agent tokens $A$ into the conventional attention module.
Smooth Diffusion: Crafting Smooth Latent Spaces in Diffusion Models
Specifically, we introduce Step-wise Variation Regularization to enforce the proportion between the variations of an arbitrary input latent and that of the output image is a constant at any diffusion training step.
Augmenting Unsupervised Reinforcement Learning with Self-Reference
Our approach achieves state-of-the-art results in terms of Interquartile Mean (IQM) performance and Optimality Gap reduction on the Unsupervised Reinforcement Learning Benchmark for model-free methods, recording an 86% IQM and a 16% Optimality Gap.
Detecting Generated Images by Real Images Only
This paper approaches the generated image detection problem from a new perspective: Start from real images.
Train Once, Get a Family: State-Adaptive Balances for Offline-to-Online Reinforcement Learning
Existing solutions tackle this problem by imposing a policy constraint on the policy improvement objective in both offline and online learning.
STORM: Efficient Stochastic Transformer based World Models for Reinforcement Learning
The performance of these algorithms heavily relies on the sequence modeling and generation capabilities of the world model.
Ranked #5 on
Atari Games 100k
on Atari 100k
Understanding, Predicting and Better Resolving Q-Value Divergence in Offline-RL
We first identify a fundamental pattern, self-excitation, as the primary cause of Q-value estimation divergence in offline RL.
Avalon's Game of Thoughts: Battle Against Deception through Recursive Contemplation
This study utilizes the intricate Avalon game as a testbed to explore LLMs' potential in deceptive environments.
Generalized Activation via Multivariate Projection
Activation functions are essential to introduce nonlinearity into neural networks, with the Rectified Linear Unit (ReLU) often favored for its simplicity and effectiveness.
Hundreds Guide Millions: Adaptive Offline Reinforcement Learning with Expert Guidance
Offline reinforcement learning (RL) optimizes the policy on a previously collected dataset without any interactions with the environment, yet usually suffers from the distributional shift problem.
DAT++: Spatially Dynamic Vision Transformer with Deformable Attention
On the one hand, using dense attention in ViT leads to excessive memory and computational cost, and features can be influenced by irrelevant parts that are beyond the region of interests.
Ranked #4 on
Object Detection
on COCO 2017
Fine-grained Recognition with Learnable Semantic Data Augmentation
Since images belonging to the same meta-category usually share similar visual appearances, mining discriminative visual cues is the key to distinguishing fine-grained categories.
Latency-aware Unified Dynamic Networks for Efficient Image Recognition
Dynamic computation has emerged as a promising avenue to enhance the inference efficiency of deep networks.
Computation-efficient Deep Learning for Computer Vision: A Survey
Over the past decade, deep learning models have exhibited considerable advancements, reaching or even exceeding human-level performance in a range of visual perception tasks.
ExpeL: LLM Agents Are Experiential Learners
The recent surge in research interest in applying large language models (LLMs) to decision-making tasks has flourished by leveraging the extensive world knowledge embedded in LLMs.
Learning Specialized Activation Functions for Physics-informed Neural Networks
To avoid the inefficient manual selection and to alleviate the optimization difficulty of PINNs, we introduce adaptive activation functions to search for the optimal function when solving different problems.
FLatten Transformer: Vision Transformer using Focused Linear Attention
The quadratic computation complexity of self-attention has been a persistent challenge when applying Transformer models to vision tasks.
Multi-Dimensional Refinement Graph Convolutional Network with Robust Decouple Loss for Fine-Grained Skeleton-Based Action Recognition
Graph convolutional networks have been widely used in skeleton-based action recognition.
Fine-grained Action Recognition
Skeleton Based Action Recognition
Dynamic Perceiver for Efficient Visual Recognition
Early exits are placed exclusively within the classification branch, thus eliminating the need for linear separability in low-level features.
Decoupled Prioritized Resampling for Offline RL
OPER is a plug-and-play component for offline RL algorithms.
ADDP: Learning General Representations for Image Recognition and Generation with Alternating Denoising Diffusion Process
In each denoising step, our method first decodes pixels from previous VQ tokens, then generates new VQ tokens from the decoded pixels.
Boosting Offline Reinforcement Learning with Action Preference Query
In particular, online fine-tuning has become a commonly used method to correct the erroneous estimates of out-of-distribution data learned in the offline training phase.
Prompt-Free Diffusion: Taking "Text" out of Text-to-Image Diffusion Models
Text-to-image (T2I) research has grown explosively in the past year, owing to the large-scale pre-trained diffusion models and many emerging personalization and editing approaches.
Ghost in the Minecraft: Generally Capable Agents for Open-World Environments via Large Language Models with Text-based Knowledge and Memory
These agents, equipped with the logic and common sense capabilities of LLMs, can skillfully navigate complex, sparse-reward environments with text-based interactions.
Slide-Transformer: Hierarchical Vision Transformer with Local Self-Attention
Self-attention mechanism has been a key factor in the recent progress of Vision Transformer (ViT), which enables adaptive feature extraction from global contexts.
Zero-shot Generative Model Adaptation via Image-specific Prompt Learning
Recently, CLIP-guided image synthesis has shown appealing performance on adapting a pre-trained source-domain generator to an unseen target domain.
Adaptive Rotated Convolution for Rotated Object Detection
In our ARC module, the convolution kernels rotate adaptively to extract object features with varying orientations in different images, and an efficient conditional computation mechanism is introduced to accommodate the large orientation variations of objects within an image.
Ranked #3 on
Object Detection In Aerial Images
on DOTA
(using extra training data)
Joint Representation Learning for Text and 3D Point Cloud
During the pre-training stage, we establish the correspondence of images and point clouds based on the readily available RGB-D data and use contrastive learning to align the image and point cloud representations.
Borrowing Knowledge From Pre-trained Language Model: A New Data-efficient Visual Learning Paradigm
To address this issue, this paper presents a novel approach that seeks to leverage linguistic knowledge for data-efficient visual learning.
Convolution-enhanced Evolving Attention Networks
To the best of our knowledge, this is the first work that explicitly models the layer-wise evolution of attention maps.
Deep Incubation: Training Large Models by Divide-and-Conquering
In this paper, we present Deep Incubation, a novel approach that enables the efficient and effective training of large models by dividing them into smaller sub-modules that can be trained separately and assembled seamlessly.
Boosted Dynamic Neural Networks
To optimize the model, these prediction heads together with the network backbone are trained on every batch of training data.
BEVFormer v2: Adapting Modern Image Backbones to Bird's-Eye-View Recognition via Perspective Supervision
The proposed method is verified with a wide spectrum of traditional and modern image backbones and achieves new SoTA results on the large-scale nuScenes dataset.
Ranked #5 on
3D Object Detection
on Rope3D
Cross-Modal Adapter for Text-Video Retrieval
However, as pre-trained models are scaling up, fully fine-tuning them on text-video retrieval datasets has a high risk of overfitting.
Towards All-in-one Pre-training via Maximizing Multi-modal Mutual Information
It has been proved that combining multiple pre-training strategies and data from various modalities/sources can greatly boost the training of large-scale models.
Ranked #2 on
Semantic Segmentation
on ADE20K
(using extra training data)
EfficientTrain: Exploring Generalized Curriculum Learning for Training Visual Backbones
It is also effective for self-supervised learning (e. g., MAE).
Boosting Offline Reinforcement Learning via Data Rebalancing
Therefore, we propose a simple yet effective method to boost offline RL algorithms based on the observation that resampling a dataset keeps the distribution support unchanged.
Contrastive Language-Image Pre-Training with Knowledge Graphs
Recent years have witnessed the fast development of large-scale pre-training frameworks that can extract multi-modal representations in a unified form and achieve promising performances when transferred to downstream tasks.
A Mixture of Surprises for Unsupervised Reinforcement Learning
However, both strategies rely on a strong assumption: the entropy of the environment's dynamics is either high or low.
Efficient Knowledge Distillation from Model Checkpoints
Knowledge distillation is an effective approach to learn compact models (students) with the supervision of large and strong models (teachers).
Latency-aware Spatial-wise Dynamic Networks
The latency prediction model can efficiently estimate the inference latency of dynamic networks by simultaneously considering algorithms, scheduling strategies, and hardware properties.
AdaFocusV3: On Unified Spatial-temporal Dynamic Video Recognition
Recent research has revealed that reducing the temporal and spatial redundancy are both effective approaches towards efficient video recognition, e. g., allocating the majority of computation to a task-relevant subset of frames or the most valuable image regions of each frame.
ActiveNeRF: Learning where to See with Uncertainty Estimation
In this paper, we present a novel learning framework, ActiveNeRF, aiming to model a 3D scene with a constrained input budget.
Learning to Weight Samples for Dynamic Early-exiting Networks
Intuitively, easy samples, which generally exit early in the network during inference, should contribute more to training early classifiers.
Value-Consistent Representation Learning for Data-Efficient Reinforcement Learning
Recently, visual representation learning has been shown to be effective and promising for boosting sample efficiency in RL.
DiSparse: Disentangled Sparsification for Multitask Model Compression
We analyzed the pruning masks generated with DiSparse and observed strikingly similar sparse network architecture identified by each task even before the training starts.
Siamese Image Modeling for Self-Supervised Vision Representation Learning
Driven by these analysis, we propose Siamese Image Modeling (SiameseIM), which predicts the dense representations of an augmented view, based on another masked view from the same image but with different augmentations.
Provable General Function Class Representation Learning in Multitask Bandits and MDPs
While multitask representation learning has become a popular approach in reinforcement learning (RL) to boost the sample efficiency, the theoretical understanding of why and how it works is still limited.
SePiCo: Semantic-Guided Pixel Contrast for Domain Adaptive Semantic Segmentation
Domain adaptive semantic segmentation attempts to make satisfactory dense predictions on an unlabeled target domain by utilizing the supervised model trained on a labeled source domain.
Ranked #4 on
Semantic Segmentation
on GTAV-to-Cityscapes Labels
Pseudo-Q: Generating Pseudo Language Queries for Visual Grounding
Our method leverages an off-the-shelf object detector to identify visual objects from unlabeled images, and then language queries for these objects are obtained in an unsupervised fashion with a pseudo-query generation module.
Learn From the Past: Experience Ensemble Knowledge Distillation
We save a moderate number of intermediate models from the training process of the teacher model uniformly, and then integrate the knowledge of these intermediate models by ensemble technique.
Domain Adaptation via Prompt Learning
Unsupervised domain adaption (UDA) aims to adapt models learned from a well-annotated source domain to a target domain, where only unlabeled samples are given.
Glance and Focus Networks for Dynamic Visual Recognition
Spatial redundancy widely exists in visual recognition tasks, i. e., discriminative features in an image or video frame usually correspond to only a subset of pixels, while the remaining regions are irrelevant to the task at hand.
Vision Transformer with Deformable Attention
On the one hand, using dense attention e. g., in ViT, leads to excessive memory and computational cost, and features can be influenced by irrelevant parts which are beyond the region of interests.
Ranked #107 on
Object Detection
on COCO test-dev
AdaFocus V2: End-to-End Training of Spatial Dynamic Networks for Video Recognition
Recent works have shown that the computational efficiency of video recognition can be significantly improved by reducing the spatial redundancy.
Exploring the Equivalence of Siamese Self-Supervised Learning via A Unified Gradient Framework
These methods appear to be quite different in the designed loss functions from various motivations.
Searching Parameterized AP Loss for Object Detection
In this paper, we propose Parameterized AP Loss, where parameterized functions are introduced to substitute the non-differentiable components in the AP calculation.
Assessing a Single Image in Reference-Guided Image Synthesis
For Reference-guided Image Synthesis (RIS) tasks, i. e., rendering a source image in the style of another reference image, where assessing the quality of a single generated image is crucial, these metrics are not applicable.
Temporal-Spatial Causal Interpretations for Vision-Based Reinforcement Learning
TSCI model builds on the formulation of temporal causality, which reflects the temporal causal relations between sequential observations and decisions of RL agent.
On the Integration of Self-Attention and Convolution
In this paper, we show that there exists a strong underlying relation between them, in the sense that the bulk of computations of these two paradigms are in fact done with the same operation.
Fine-Grained Few Shot Learning with Foreground Object Transformation
As a data augmentation method, FOT can be conveniently applied to any existing few shot learning algorithm and greatly improve its performance on FG-FSL tasks.
CAM-loss: Towards Learning Spatially Discriminative Feature Representations
The backbone of traditional CNN classifier is generally considered as a feature extractor, followed by a linear layer which performs the classification.
Integrating Large Circular Kernels into CNNs through Neural Architecture Search
We first propose a simple yet efficient implementation of the convolution using circular kernels, and empirically show the significant advantages of large circular kernels over the counterpart square kernels.
On the Power of Multitask Representation Learning in Linear MDP
We first discover a \emph{Least-Activated-Feature-Abundance} (LAFA) criterion, denoted as $\kappa$, with which we prove that a straightforward least-square algorithm learns a policy which is $\tilde{O}(H^2\sqrt{\frac{\mathcal{C}(\Phi)^2 \kappa d}{NT}+\frac{\kappa d}{n}})$ sub-optimal.
Believe What You See: Implicit Constraint Approach for Offline Multi-Agent Reinforcement Learning
Moreover, we extend ICQ to multi-agent tasks by decomposing the joint-policy under the implicit constraint.
Not All Images are Worth 16x16 Words: Dynamic Transformers for Efficient Image Recognition
Inspired by this phenomenon, we propose a Dynamic Transformer to automatically configure a proper number of tokens for each input image.
Ranked #29 on
Image Classification
on CIFAR-100
(using extra training data)
Adaptive Focus for Efficient Video Recognition
In this paper, we explore the spatial redundancy in video recognition with the aim to improve the computational efficiency.
CondenseNet V2: Sparse Feature Reactivation for Deep Networks
Reusing features in deep networks through dense connectivity is an effective way to achieve high computational efficiency.
AutoLoss-Zero: Searching Loss Functions from Scratch for Generic Tasks
However, the automatic design of loss functions for generic tasks with various evaluation metrics remains under-investigated.
Generalized Domain Conditioned Adaptation Network
Domain Adaptation (DA) attempts to transfer knowledge learned in the labeled source domain to the unlabeled but related target domain without requiring large amounts of target supervision.
Evolving Attention with Residual Convolutions
In this paper, we propose a novel and generic mechanism based on evolving attention to improve the performance of transformers.
Dynamic Neural Networks: A Survey
Dynamic neural network is an emerging research topic in deep learning.
Revisiting Locally Supervised Learning: an Alternative to End-to-end Training
Due to the need to store the intermediate activations for back-propagation, end-to-end (E2E) training of deep networks usually suffers from high GPUs memory footprint.
Revisiting Locally Supervised Training of Deep Neural Networks
As InfoPro loss is difficult to compute in its original form, we derive a feasible upper bound as a surrogate optimization objective, yielding a simple but effective algorithm.
Robust Offline Reinforcement Learning from Low-Quality Data
We theoretically show that AdaPT produces a tight upper bound on the distributional deviation between the learned policy and the behavior policy, and this upper bound is the minimum requirement to guarantee policy improvement at each iteration.
A Unified Framework for Convolution-based Graph Neural Networks
In this paper, we take a step forward to establish a unified framework for convolution-based graph neural networks, by formulating the basic graph convolution operation as an optimization problem in the graph Fourier space.
3D Object Detection with Pointformer
In this paper, we propose Pointformer, a Transformer backbone designed for 3D point clouds to learn features effectively.
Frequency Domain Image Translation: More Photo-realistic, Better Identity-preserving
Image-to-image translation has been revolutionized with GAN-based methods.
Auto Seg-Loss: Searching Metric Surrogates for Semantic Segmentation
In this paper, we propose to automate the design of metric-specific loss functions by searching differentiable surrogate losses for each metric.
Glance and Focus: a Dynamic Approach to Reducing Spatial Redundancy in Image Classification
The accuracy of deep convolutional neural networks (CNNs) generally improves when fueled with high resolution images.
Regularizing Deep Networks with Semantic Data Augmentation
The proposed method is inspired by the intriguing property that deep networks are effective in learning linearized features, i. e., certain directions in the deep feature space correspond to meaningful semantic transformations, e. g., changing the background or view angle of an object.
Meta-Semi: A Meta-learning Approach for Semi-supervised Learning
In this paper, we propose a novel meta-learning based SSL algorithm (Meta-Semi) that requires tuning only one additional hyper-parameter, compared with a standard supervised deep learning algorithm, to achieve competitive performance under various conditions of SSL.
Domain Conditioned Adaptation Network
Most existing deep DA models only focus on aligning feature representations of task-specific layers across domains while integrating a totally shared convolutional architecture for source and target.
Deep Residual Correction Network for Partial Domain Adaptation
Deep domain adaptation methods have achieved appealing performance by learning transferable representations from a well-labeled source domain to a different but related unlabeled target domain.
Spatially Adaptive Inference with Stochastic Feature Sampling and Interpolation
In the feature maps of CNNs, there commonly exists considerable spatial redundancy that leads to much repetitive processing.
Resolution Adaptive Networks for Efficient Inference
Adaptive inference is an effective mechanism to achieve a dynamic tradeoff between accuracy and computational cost in deep networks.
Self-Supervised Discovering of Interpretable Features for Reinforcement Learning
Deep reinforcement learning (RL) has recently led to many breakthroughs on a range of complex control tasks.
Tighter Bound Estimation of Sensitivity Analysis for Incremental and Decremental Data Modification
Specifically, the proposed algorithm can be used to estimate the upper and lower bounds of the updated classifier's coefficient matrix with a low computational complexity related to the size of the updated dataset.
Cross-Iteration Batch Normalization
We thus compensate for the network weight changes via a proposed technique based on Taylor polynomials, so that the statistics can be accurately estimated and batch normalization can be effectively applied.
Ranked #180 on
Object Detection
on COCO test-dev
FSD-10: A Dataset for Competitive Sports Content Analysis
To promote the research on action recognition from competitive sports video clips, we introduce a Figure Skating Dataset (FSD-10) for finegrained sports content analysis.
Gated Path Selection Network for Semantic Segmentation
Semantic segmentation is a challenging task that needs to handle large scale variations, deformations and different viewpoints.
Convolutional Networks with Dense Connectivity
Recent work has shown that convolutional networks can be substantially deeper, more accurate, and efficient to train if they contain shorter connections between layers close to the input and those close to the output.
Implicit Semantic Data Augmentation for Deep Networks
Our work is motivated by the intriguing property that deep networks are surprisingly good at linearizing features, such that certain directions in the deep feature space correspond to meaningful semantic transformations, e. g., adding sunglasses or changing backgrounds.
Regularized Anderson Acceleration for Off-Policy Deep Reinforcement Learning
To tackle this problem, we propose a general acceleration method for model-free, off-policy deep RL algorithms by drawing the idea underlying regularized Anderson acceleration (RAA), which is an effective approach to accelerating the solving of fixed point problems with perturbations.
Improved Techniques for Training Adaptive Deep Networks
Adaptive inference is a promising technique to improve the computational efficiency of deep models at test time.
Asymmetric Valleys: Beyond Sharp and Flat Local Minima
Specifically, at a local minimum there exist many asymmetric directions such that the loss increases abruptly along one side, and slowly along the opposite side--we formally define such minima as asymmetric valleys.
Gradient Boosted Feature Selection
A feature selection algorithm should ideally satisfy four conditions: reliably extract relevant features; be able to identify non-linear feature interactions; scale linearly with the number of features and dimensions; allow the incorporation of known sparsity structure.
Domain-Aware SE Network for Sketch-based Image Retrieval with Multiplicative Euclidean Margin Softmax
This paper proposes a novel approach for Sketch-Based Image Retrieval (SBIR), for which the key is to bridge the gap between sketches and photos in terms of the data representation.
Anytime Stereo Image Depth Estimation on Mobile Devices
Many applications of stereo depth estimation in robotics require the generation of accurate disparity maps in real time under significant computational constraints.
Ranked #1 on
Stereo Depth Estimation
on KITTI2012
Rethinking the Value of Network Pruning
Our observations are consistent for multiple network architectures, datasets, and tasks, which imply that: 1) training a large, over-parameterized model is often not necessary to obtain an efficient final model, 2) learned "important" weights of the large model are typically not useful for the small pruned model, 3) the pruned architecture itself, rather than a set of inherited "important" weights, is more crucial to the efficiency in the final model, which suggests that in some cases pruning can be useful as an architecture search paradigm.
Interpretable Spatio-temporal Attention for Video Action Recognition
Inspired by the observation that humans are able to process videos efficiently by only paying attention where and when it is needed, we propose an interpretable and easy plug-in spatial-temporal attention mechanism for video action recognition.
An empirical study on evaluation metrics of generative adversarial networks
Evaluating generative adversarial networks (GANs) is inherently challenging.
Resource Aware Person Re-identification across Multiple Resolutions
Not all people are equally easy to identify: color statistics might be enough for some cases while others might require careful reasoning about high- and low-level details.
Ranked #12 on
Person Re-Identification
on CUHK03 detected
Horizontal Pyramid Matching for Person Re-identification
Despite the remarkable recent progress, person re-identification (Re-ID) approaches are still suffering from the failure cases where the discriminative body parts are missing.
Ranked #55 on
Person Re-Identification
on DukeMTMC-reID
CondenseNet: An Efficient DenseNet using Learned Group Convolutions
It combines dense connectivity with a novel module called learned group convolution.
Learning Efficient Convolutional Networks through Network Slimming
For VGGNet, a multi-pass version of network slimming gives a 20x reduction in model size and a 5x reduction in computing operations.
Memory-Efficient Implementation of DenseNets
A 264-layer DenseNet (73M parameters), which previously would have been infeasible to train, can now be trained on a single workstation with 8 NVIDIA Tesla M40 GPUs.
Snapshot Ensembles: Train 1, get M for free
In this paper, we propose a method to obtain the seemingly contradictory goal of ensembling multiple neural networks at no additional training cost.
Multi-Scale Dense Networks for Resource Efficient Image Classification
In this paper we investigate image classification with computational resource limits at test time.
Supervised Word Mover's Distance
Accurately measuring the similarity between text documents lies at the core of many real world applications of machine learning.
Densely Connected Convolutional Networks
Recent work has shown that convolutional networks can be substantially deeper, more accurate, and efficient to train if they contain shorter connections between layers close to the input and those close to the output.
Ranked #1 on
Classification
on XImageNet-12
Deep Networks with Stochastic Depth
With stochastic depth we can increase the depth of residual networks even beyond 1200 layers and still yield meaningful improvements in test error (4. 91% on CIFAR-10).
Ranked #21 on
Image Classification
on SVHN
