Search Results for author:
Found 199 papers, 79 papers with code
Dual Risk Minimization: Towards Next-Level Robustness in Fine-tuning Zero-Shot Models
Fine-tuning foundation models often compromises their robustness to distribution shifts.
Efficient Multi-modal Large Language Models via Visual Token Grouping
The development of Multi-modal Large Language Models (MLLMs) enhances Large Language Models (LLMs) with the ability to perceive data formats beyond text, significantly advancing a range of downstream applications, such as visual question answering and image captioning.
MagicDriveDiT: High-Resolution Long Video Generation for Autonomous Driving with Adaptive Control
The rapid advancement of diffusion models has greatly improved video synthesis, especially in controllable video generation, which is essential for applications like autonomous driving.
Forewarned is Forearmed: Leveraging LLMs for Data Synthesis through Failure-Inducing Exploration
In this paper, we present a novel approach, ReverseGen, designed to automatically generate effective training samples that expose the weaknesses of LLMs.
Beyond Autoregression: Discrete Diffusion for Complex Reasoning and Planning
Our work highlights the potential of diffusion-based approaches in advancing AI capabilities for sophisticated language understanding and problem-solving tasks.
How Numerical Precision Affects Mathematical Reasoning Capabilities of LLMs
Despite the remarkable success of Transformer-based Large Language Models (LLMs) across various domains, understanding and enhancing their mathematical capabilities remains a significant challenge.
DAPE V2: Process Attention Score as Feature Map for Length Extrapolation
The attention mechanism is a fundamental component of the Transformer model, contributing to interactions among distinct tokens, in contrast to earlier feed-forward neural networks.
Accelerating Auto-regressive Text-to-Image Generation with Training-free Speculative Jacobi Decoding
In this paper, we propose a training-free probabilistic parallel decoding algorithm, Speculative Jacobi Decoding (SJD), to accelerate auto-regressive text-to-image generation.
EMOVA: Empowering Language Models to See, Hear and Speak with Vivid Emotions
GPT-4o, an omni-modal model that enables vocal conversations with diverse emotions and tones, marks a milestone for omni-modal foundation models.
CoCA: Regaining Safety-awareness of Multimodal Large Language Models with Constitutional Calibration
The deployment of multimodal large language models (MLLMs) has demonstrated remarkable success in engaging in conversations involving visual inputs, thanks to the superior power of large language models (LLMs).
T2V-CompBench: A Comprehensive Benchmark for Compositional Text-to-video Generation
In this work, we conduct the first systematic study on compositional text-to-video generation.
GenArtist: Multimodal LLM as an Agent for Unified Image Generation and Editing
For a complex problem, the MLLM agent decomposes it into simpler sub-problems and constructs a tree structure to systematically plan the procedure of generation, editing, and self-correction with step-by-step verification.
Jailbreaking as a Reward Misspecification Problem
In this paper, we propose a novel perspective that attributes this vulnerability to reward misspecification during the alignment process.
QuickLLaMA: Query-aware Inference Acceleration for Large Language Models
On widely recognized benchmarks, Q-LLM improved by 7. 17% compared to the current state-of-the-art on LLaMA3, and by 3. 26% on Mistral on the $\infty$-bench.
Your Absorbing Discrete Diffusion Secretly Models the Conditional Distributions of Clean Data
In this paper, we reveal that the concrete score in absorbing diffusion can be expressed as conditional probabilities of clean data, multiplied by a time-dependent scalar in an analytic form.
The Buffer Mechanism for Multi-Step Information Reasoning in Language Models
Large language models have consistently struggled with complex reasoning tasks, such as mathematical problem-solving.
Towards Understanding the Working Mechanism of Text-to-Image Diffusion Model
We conclude that in the earlier generation stage, the image is mostly decided by the special token [\texttt{EOS}] in the text prompt, and the information in the text prompt is already conveyed in this stage.
Enhancing Text-to-Image Editing via Hybrid Mask-Informed Fusion
Then following a fusion process that carefully integrates the source intermediate (hidden) states (obtained by inversion) with the ones of the target image.
Proving Theorems Recursively
This approach allows the theorem to be tackled incrementally by outlining the overall theorem at the first level and then solving the intermediate conjectures at deeper levels.
MagicDrive3D: Controllable 3D Generation for Any-View Rendering in Street Scenes
While controllable generative models for images and videos have achieved remarkable success, high-quality models for 3D scenes, particularly in unbounded scenarios like autonomous driving, remain underdeveloped due to high data acquisition costs.
DAPE: Data-Adaptive Positional Encoding for Length Extrapolation
Positional encoding plays a crucial role in transformers, significantly impacting model performance and length generalization.
DiM: Diffusion Mamba for Efficient High-Resolution Image Synthesis
In addition, to further improve training efficiency for high-resolution image generation with DiM, we investigate "weak-to-strong" training strategy that pretrains DiM on low-resolution images ($256\times 256$) and then finetune it on high-resolution images ($512 \times 512$).
ATG: Benchmarking Automated Theorem Generation for Generative Language Models
We conduct extensive experiments to investigate whether current LMs can generate theorems in the library and benefit the problem theorems proving.
Mixture of insighTful Experts (MoTE): The Synergy of Thought Chains and Expert Mixtures in Self-Alignment
As the capabilities of large language models (LLMs) have expanded dramatically, aligning these models with human values presents a significant challenge.
Automated Evaluation of Large Vision-Language Models on Self-driving Corner Cases
Moreover, with our CODA-LM, we build CODA-VLM, a new driving LVLM surpassing all open-sourced counterparts on CODA-LM.
DetCLIPv3: Towards Versatile Generative Open-vocabulary Object Detection
This is followed by a fine-tuning stage that leverages a small number of high-resolution samples to further enhance detection performance.
Ranked #2 on
Object Detection
on ODinW Full-Shot 13 Tasks
DriveCoT: Integrating Chain-of-Thought Reasoning with End-to-End Driving
We utilize the challenging driving scenarios from the CARLA leaderboard 2. 0, which involve high-speed driving and lane-changing, and propose a rule-based expert policy to control the vehicle and generate ground truth labels for its reasoning process across different driving aspects and the final decisions.
CVT-xRF: Contrastive In-Voxel Transformer for 3D Consistent Radiance Fields from Sparse Inputs
While existing works mainly consider ray-level consistency to construct 2D learning regularization based on rendered color, depth, or semantics on image planes, in this paper we propose a novel approach that models 3D spatial field consistency to improve NeRF's performance with sparse inputs.
DetDiffusion: Synergizing Generative and Perceptive Models for Enhanced Data Generation and Perception
Furthermore, image syntheses from DetDiffusion can effectively augment training data, significantly enhancing downstream detection performance.
Editing Massive Concepts in Text-to-Image Diffusion Models
We further propose a comprehensive benchmark, named ImageNet Concept Editing Benchmark (ICEB), for evaluating massive concept editing for T2I models with two subtasks, free-form prompts, massive concept categories, and extensive evaluation metrics.
Efficient Transferability Assessment for Selection of Pre-trained Detectors
Given this zoo, we adopt 7 target datasets from 5 diverse domains as the downstream target tasks for evaluation.
Open-Vocabulary Object Detection with Meta Prompt Representation and Instance Contrastive Optimization
Without using knowledge distillation, ensemble model or extra training data during detector training, our proposed MIC outperforms previous SOTA methods trained with these complex techniques on LVIS.
Eyes Closed, Safety On: Protecting Multimodal LLMs via Image-to-Text Transformation
To construct robust MLLMs, we propose ECSO (Eyes Closed, Safety On), a novel training-free protecting approach that exploits the inherent safety awareness of MLLMs, and generates safer responses via adaptively transforming unsafe images into texts to activate the intrinsic safety mechanism of pre-aligned LLMs in MLLMs.
PixArt-Σ: Weak-to-Strong Training of Diffusion Transformer for 4K Text-to-Image Generation
In this paper, we introduce PixArt-\Sigma, a Diffusion Transformer model~(DiT) capable of directly generating images at 4K resolution.
Accelerating Diffusion Sampling with Optimized Time Steps
While this is a significant development, most sampling methods still employ uniform time steps, which is not optimal when using a small number of steps.
The Surprising Effectiveness of Skip-Tuning in Diffusion Sampling
Surprisingly, the improvement persists when we increase the number of sampling steps and can even surpass the best result from EDM-2 (1. 58) with only 39 NFEs (1. 57).
On the Expressive Power of a Variant of the Looped Transformer
Previous works try to explain this from the expressive power and capability perspectives that standard transformers are capable of performing some algorithms.
MUSTARD: Mastering Uniform Synthesis of Theorem and Proof Data
Recent large language models (LLMs) have witnessed significant advancement in various tasks, including mathematical reasoning and theorem proving.
Diffusion of Thoughts: Chain-of-Thought Reasoning in Diffusion Language Models
Recently, diffusion models have garnered significant interest in the field of text processing due to their many potential advantages compared to conventional autoregressive models.
Task-customized Masked AutoEncoder via Mixture of Cluster-conditional Experts
It also obtains new state-of-the-art self-supervised learning results on detection and segmentation.
Divide and Conquer: Language Models can Plan and Self-Correct for Compositional Text-to-Image Generation
Given a complex text prompt containing multiple concepts including objects, attributes, and relationships, the LLM agent initially decomposes it, which entails the extraction of individual objects, their associated attributes, and the prediction of a coherent scene layout.
CustomVideo: Customizing Text-to-Video Generation with Multiple Subjects
Customized text-to-video generation aims to generate high-quality videos guided by text prompts and subject references.
PIXART-δ: Fast and Controllable Image Generation with Latent Consistency Models
As a state-of-the-art, open-source image generation model, PIXART-{\delta} offers a promising alternative to the Stable Diffusion family of models, contributing significantly to text-to-image synthesis.
Enhancing the Power of OOD Detection via Sample-Aware Model Selection
In this work we present a novel perspective on detecting out-of-distribution (OOD) samples and propose an algorithm for sample-aware model selection to enhance the effectiveness of OOD detection.
SERF: Fine-Grained Interactive 3D Segmentation and Editing with Radiance Fields
Although significant progress has been made in the field of 2D-based interactive editing, fine-grained 3D-based interactive editing remains relatively unexplored.
G-LLaVA: Solving Geometric Problem with Multi-Modal Large Language Model
We first analyze the limitations of current Multimodal Large Language Models (MLLMs) in this area: they struggle to accurately comprehending basic geometric elements and their relationships.
A Survey of Reasoning with Foundation Models
Reasoning, a crucial ability for complex problem-solving, plays a pivotal role in various real-world settings such as negotiation, medical diagnosis, and criminal investigation.
Fast Training of Diffusion Transformer with Extreme Masking for 3D Point Clouds Generation
Motivated by the inherent redundancy of 3D compared to 2D, we propose FastDiT-3D, a novel masked diffusion transformer tailored for efficient 3D point cloud generation, which greatly reduces training costs.
Drag-A-Video: Non-rigid Video Editing with Point-based Interaction
In this paper, we propose a new diffusion-based method for interactive point-based video manipulation, called Drag-A-Video.
Animate124: Animating One Image to 4D Dynamic Scene
We introduce Animate124 (Animate-one-image-to-4D), the first work to animate a single in-the-wild image into 3D video through textual motion descriptions, an underexplored problem with significant applications.
Gaining Wisdom from Setbacks: Aligning Large Language Models via Mistake Analysis
The rapid development of large language models (LLMs) has not only provided numerous opportunities but also presented significant challenges.
Robustness May be More Brittle than We Think under Different Degrees of Distribution Shifts
This indicates that while pre-trained representations may help improve downstream in-distribution performance, they could have minimal or even adverse effects on generalization in certain OOD scenarios of the downstream task if not used properly.
Implicit Concept Removal of Diffusion Models
Subsequently, the model is optimized to identify and disentangle this information, which is then adopted as negative prompts during generation.
MagicDrive: Street View Generation with Diverse 3D Geometry Control
Recent advancements in diffusion models have significantly enhanced the data synthesis with 2D control.
BYOM: Building Your Own Multi-Task Model For Free
Recently, various merging methods have been proposed to build a multi-task model from task-specific finetuned models without retraining.
DriveGPT4: Interpretable End-to-end Autonomous Driving via Large Language Model
Multimodal large language models (MLLMs) have emerged as a prominent area of interest within the research community, given their proficiency in handling and reasoning with non-textual data, including images and videos.
LEGO-Prover: Neural Theorem Proving with Growing Libraries
Our ablation study indicates that these newly added skills are indeed helpful for proving theorems, resulting in an improvement from a success rate of 47. 1% to 50. 4%.
Ranked #1 on
Automated Theorem Proving
on miniF2F-test
(Pass@100 metric)
PixArt-$α$: Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesis
We hope PIXART-$\alpha$ will provide new insights to the AIGC community and startups to accelerate building their own high-quality yet low-cost generative models from scratch.
Lyra: Orchestrating Dual Correction in Automated Theorem Proving
In addition, we introduce Conjecture Correction, an error feedback mechanism designed to interact with prover to refine formal proof conjectures with prover error messages.
Ranked #1 on
Automated Theorem Proving
on miniF2F-test
(Pass@100 metric)
MetaMath: Bootstrap Your Own Mathematical Questions for Large Language Models
Our MetaMath-7B model achieves 66. 4% on GSM8K and 19. 4% on MATH, exceeding the state-of-the-art models of the same size by 11. 5% and 8. 7%.
Ranked #57 on
Arithmetic Reasoning
on GSM8K
(using extra training data)
SA-Solver: Stochastic Adams Solver for Fast Sampling of Diffusion Models
Based on our analysis, we propose SA-Solver, which is an improved efficient stochastic Adams method for solving diffusion SDE to generate data with high quality.
Ranked #20 on
Image Generation
on ImageNet 512x512
Beyond One-to-One: Rethinking the Referring Image Segmentation
In this paper, we address this issue from two perspectives.
Forward-Backward Reasoning in Large Language Models for Mathematical Verification
Instead of using forward or backward reasoning alone, we propose FOBAR to combine FOrward and BAckward Reasoning for verification.
UniTR: A Unified and Efficient Multi-Modal Transformer for Bird's-Eye-View Representation
Jointly processing information from multiple sensors is crucial to achieving accurate and robust perception for reliable autonomous driving systems.
Ranked #8 on
3D Object Detection
on nuScenes
A Causal Framework to Unify Common Domain Generalization Approaches
Domain generalization (DG) is about learning models that generalize well to new domains that are related to, but different from, the training domain(s).
T2I-CompBench: A Comprehensive Benchmark for Open-world Compositional Text-to-image Generation
Despite the stunning ability to generate high-quality images by recent text-to-image models, current approaches often struggle to effectively compose objects with different attributes and relationships into a complex and coherent scene.
DiffFlow: A Unified SDE Framework for Score-Based Diffusion Models and Generative Adversarial Networks
Generative models can be categorized into two types: explicit generative models that define explicit density forms and allow exact likelihood inference, such as score-based diffusion models (SDMs) and normalizing flows; implicit generative models that directly learn a transformation from the prior to the data distribution, such as generative adversarial nets (GANs).
Training Energy-Based Models with Diffusion Contrastive Divergences
In image generation experiments, the proposed DCD is capable of training an energy-based model for generating the Celab-A $32\times 32$ dataset, which is comparable to existing EBMs.
DiT-3D: Exploring Plain Diffusion Transformers for 3D Shape Generation
Recent Diffusion Transformers (e. g., DiT) have demonstrated their powerful effectiveness in generating high-quality 2D images.
Ranked #1 on
Point Cloud Generation
on ShapeNet Car
GeoDiffusion: Text-Prompted Geometric Control for Object Detection Data Generation
Diffusion models have attracted significant attention due to the remarkable ability to create content and generate data for tasks like image classification.
Explore and Exploit the Diverse Knowledge in Model Zoo for Domain Generalization
The proliferation of pretrained models, as a result of advancements in pretraining techniques, has led to the emergence of a vast zoo of publicly available models.
Diff-Instruct: A Universal Approach for Transferring Knowledge From Pre-trained Diffusion Models
To demonstrate the effectiveness and universality of Diff-Instruct, we consider two scenarios: distilling pre-trained diffusion models and refining existing GAN models.
On the Generalization of Diffusion Model
To understand this contradiction, we empirically verify the difference between the sufficiently trained diffusion model and the empirical optima.
ConsistentNeRF: Enhancing Neural Radiance Fields with 3D Consistency for Sparse View Synthesis
In this paper, we propose ConsistentNeRF, a method that leverages depth information to regularize both multi-view and single-view 3D consistency among pixels.
Make-A-Protagonist: Generic Video Editing with An Ensemble of Experts
The text-driven image and video diffusion models have achieved unprecedented success in generating realistic and diverse content.
Boosting Visual-Language Models by Exploiting Hard Samples
In this work, we present HELIP, a cost-effective strategy tailored to enhance the performance of existing CLIP models without the need for training a model from scratch or collecting additional data.
MetaBEV: Solving Sensor Failures for BEV Detection and Map Segmentation
These queries are then processed iteratively by a BEV-Evolving decoder, which selectively aggregates deep features from either LiDAR, cameras, or both modalities.
Progressive-Hint Prompting Improves Reasoning in Large Language Models
The performance of Large Language Models (LLMs) in reasoning tasks depends heavily on prompt design, with Chain-of-Thought (CoT) and self-consistency being critical methods that enhance this ability.
Ranked #4 on
Math Word Problem Solving
on SVAMP
DiffFit: Unlocking Transferability of Large Diffusion Models via Simple Parameter-Efficient Fine-Tuning
This paper proposes DiffFit, a parameter-efficient strategy to fine-tune large pre-trained diffusion models that enable fast adaptation to new domains.
Efficient Diffusion Personalization
parameter-efficient fine-tuning
DetCLIPv2: Scalable Open-Vocabulary Object Detection Pre-training via Word-Region Alignment
This paper presents DetCLIPv2, an efficient and scalable training framework that incorporates large-scale image-text pairs to achieve open-vocabulary object detection (OVD).
Ranked #5 on
Object Detection
on ODinW Full-Shot 13 Tasks
DeepAccident: A Motion and Accident Prediction Benchmark for V2X Autonomous Driving
In addition, we propose a new task, end-to-end motion and accident prediction, which can be used to directly evaluate the accident prediction ability for different autonomous driving algorithms.
Fair-CDA: Continuous and Directional Augmentation for Group Fairness
By adjusting the perturbation strength in the direction of the paths, our proposed augmentation is controllable and auditable.
Mixed Autoencoder for Self-supervised Visual Representation Learning
Specifically, our MixedAE outperforms MAE by +0. 3% accuracy, +1. 7 mIoU and +0. 9 AP on ImageNet-1K, ADE20K and COCO respectively with a standard ViT-Base.
DDP: Diffusion Model for Dense Visual Prediction
We propose a simple, efficient, yet powerful framework for dense visual predictions based on the conditional diffusion pipeline.
Ranked #3 on
Monocular Depth Estimation
on SUN-RGBD
ContraNeRF: Generalizable Neural Radiance Fields for Synthetic-to-real Novel View Synthesis via Contrastive Learning
In this work, we first investigate the effects of synthetic data in synthetic-to-real novel view synthesis and surprisingly observe that models trained with synthetic data tend to produce sharper but less accurate volume densities.
Entity-Level Text-Guided Image Manipulation
In the image manipulation phase, SeMani adopts a generative model to synthesize new images conditioned on the entity-irrelevant regions and target text descriptions.
MetaBEV: Solving Sensor Failures for 3D Detection and Map Segmentation
These queries are then processed iteratively by a BEV-Evolving decoder, which selectively aggregates deep features from either LiDAR, cameras, or both modalities.
Boosting Out-of-Distribution Detection with Multiple Pre-trained Models
This advance raises a natural question: Can we leverage the diversity of multiple pre-trained models to improve the performance of post hoc detection methods?
Generative Negative Text Replay for Continual Vision-Language Pretraining
In this work, we focus on learning a VLP model with sequential chunks of image-text pair data.
Dual-Curriculum Teacher for Domain-Inconsistent Object Detection in Autonomous Driving
Specifically, DucTeacher consists of two curriculums, i. e., (1) domain evolving curriculum seeks to learn from the data progressively to handle data distribution discrepancy by estimating the similarity between domains, and (2) distribution matching curriculum seeks to estimate the class distribution for each unlabeled domain to handle class distribution shifts.
ZooD: Exploiting Model Zoo for Out-of-Distribution Generalization
We evaluate our paradigm on a diverse model zoo consisting of 35 models for various OoD tasks and demonstrate: (i) model ranking is better correlated with fine-tuning ranking than previous methods and up to 9859x faster than brute-force fine-tuning; (ii) OoD generalization after model ensemble with feature selection outperforms the state-of-the-art methods and the accuracy on most challenging task DomainNet is improved from 46. 5\% to 50. 6\%.
CAGroup3D: Class-Aware Grouping for 3D Object Detection on Point Clouds
We present a novel two-stage fully sparse convolutional 3D object detection framework, named CAGroup3D.
Ranked #1 on
3D Object Detection
on SUN-RGBD
DetCLIP: Dictionary-Enriched Visual-Concept Paralleled Pre-training for Open-world Detection
We further design a concept dictionary~(with descriptions) from various online sources and detection datasets to provide prior knowledge for each concept.
DevNet: Self-supervised Monocular Depth Learning via Density Volume Construction
Self-supervised depth learning from monocular images normally relies on the 2D pixel-wise photometric relation between temporally adjacent image frames.
Learning to Prove Trigonometric Identities
We define the normalized form of trigonometric identities, design a set of rules for the proof and put forward a method which can generate theoretically infinite trigonometric identities.
Breaking Correlation Shift via Conditional Invariant Regularizer
The correlation shift is caused by the spurious attributes that correlate to the class label, as the correlation between them may vary in training and test data.
PILC: Practical Image Lossless Compression with an End-to-end GPU Oriented Neural Framework
Generative model based image lossless compression algorithms have seen a great success in improving compression ratio.
CO^3: Cooperative Unsupervised 3D Representation Learning for Autonomous Driving
In this paper, we propose CO^3, namely Cooperative Contrastive Learning and Contextual Shape Prediction, to learn 3D representation for outdoor-scene point clouds in an unsupervised manner.
Task-Customized Self-Supervised Pre-training with Scalable Dynamic Routing
On the other hand, for existing SSL methods, it is burdensome and infeasible to use different downstream-task-customized datasets in pre-training for different tasks.
Self-Guided Noise-Free Data Generation for Efficient Zero-Shot Learning
In this paradigm, the synthesized data from the PLM acts as the carrier of knowledge, which is used to train a task-specific model with orders of magnitude fewer parameters than the PLM, achieving both higher performance and efficiency than prompt-based zero-shot learning methods on PLMs.
Arch-Graph: Acyclic Architecture Relation Predictor for Task-Transferable Neural Architecture Search
It is able to find top 0. 16\% and 0. 29\% architectures on average on two search spaces under the budget of only 50 models.
ManiTrans: Entity-Level Text-Guided Image Manipulation via Token-wise Semantic Alignment and Generation
Existing text-guided image manipulation methods aim to modify the appearance of the image or to edit a few objects in a virtual or simple scenario, which is far from practical application.
Generalizing Few-Shot NAS with Gradient Matching
Efficient performance estimation of architectures drawn from large search spaces is essential to Neural Architecture Search.
On Redundancy and Diversity in Cell-based Neural Architecture Search
Searching for the architecture cells is a dominant paradigm in NAS.
CODA: A Real-World Road Corner Case Dataset for Object Detection in Autonomous Driving
One main reason that impedes the development of truly reliably self-driving systems is the lack of public datasets for evaluating the performance of object detectors on corner cases.
Revisiting Over-smoothing in BERT from the Perspective of Graph
Recently over-smoothing phenomenon of Transformer-based models is observed in both vision and language fields.
Memory Replay with Data Compression for Continual Learning
In this work, we propose memory replay with data compression (MRDC) to reduce the storage cost of old training samples and thus increase their amount that can be stored in the memory buffer.
DropNAS: Grouped Operation Dropout for Differentiable Architecture Search
Neural architecture search (NAS) has shown encouraging results in automating the architecture design.
Semi-Supervised Object Detection via Multi-Instance Alignment With Global Class Prototypes
Semi-Supervised object detection (SSOD) aims to improve the generalization ability of object detectors with large-scale unlabeled images.
Long-tail Recognition via Compositional Knowledge Transfer
In this work, we introduce a novel strategy for long-tail recognition that addresses the tail classes' few-shot problem via training-free knowledge transfer.
Layer-Parallel Training of Residual Networks with Auxiliary-Variable Networks
Gradient-based methods for the distributed training of residual networks (ResNets) typically require a forward pass of the input data, followed by back-propagating the error gradient to update model parameters, which becomes time-consuming as the network goes deeper.
Understanding Square Loss in Training Overparametrized Neural Network Classifiers
Comparing to cross-entropy, square loss has comparable generalization error but noticeable advantages in robustness and model calibration.
On Effective Scheduling of Model-based Reinforcement Learning
Model-based reinforcement learning has attracted wide attention due to its superior sample efficiency.
MixACM: Mixup-Based Robustness Transfer via Distillation of Activated Channel Maps
First, we theoretically show the transferability of robustness from an adversarially trained teacher model to a student model with the help of mixup augmentation.
FILIP: Fine-grained Interactive Language-Image Pre-Training
In this paper, we introduce a large-scale Fine-grained Interactive Language-Image Pre-training (FILIP) to achieve finer-level alignment through a cross-modal late interaction mechanism, which uses a token-wise maximum similarity between visual and textual tokens to guide the contrastive objective.
AIM: Automatic Interaction Machine for Click-Through Rate Prediction
To address these three issues mentioned above, we propose Automatic Interaction Machine (AIM) with three core components, namely, Feature Interaction Search (FIS), Interaction Function Search (IFS) and Embedding Dimension Search (EDS), to select significant feature interactions, appropriate interaction functions and necessary embedding dimensions automatically in a unified framework.
OSOA: One-Shot Online Adaptation of Deep Generative Models for Lossless Compression
To eliminate the requirement of saving separate models for different target datasets, we propose a novel setting that starts from a pretrained deep generative model and compresses the data batches while adapting the model with a dynamical system for only one epoch.
iFlow: Numerically Invertible Flows for Efficient Lossless Compression via a Uniform Coder
In this paper, we discuss lossless compression using normalizing flows which have demonstrated a great capacity for achieving high compression ratios.
Ranked #1 on
Image Compression
on ImageNet32
Rethinking Adversarial Transferability from a Data Distribution Perspective
Adversarial transferability enables attackers to generate adversarial examples from the source model to attack the target model, which has raised security concerns about the deployment of DNNs in practice.
Nonlinear ICA Using Volume-Preserving Transformations
Nonlinear ICA is a fundamental problem in machine learning, aiming to identify the underlying independent components (sources) from data which is assumed to be a nonlinear function (mixing function) of these sources.
How Well Does Self-Supervised Pre-Training Perform with Streaming ImageNet?
Prior works on self-supervised pre-training focus on the joint training scenario, where massive unlabeled data are assumed to be given as input all at once, and only then is a learner trained.
Layer-Parallel Training of Residual Networks with Auxiliary Variables
Backpropagation algorithm is indispensable for training modern residual networks (ResNets) and usually tends to be time-consuming due to its inherent algorithmic lockings.
DHA: End-to-End Joint Optimization of Data Augmentation Policy, Hyper-parameter and Architecture
In view of these, we propose DHA, which achieves joint optimization of Data augmentation policy, Hyper-parameter and Architecture.
NAS-OoD: Neural Architecture Search for Out-of-Distribution Generalization
In this work, we propose robust Neural Architecture Search for OoD generalization (NAS-OoD), which optimizes the architecture with respect to its performance on generated OoD data by gradient descent.
Ranked #1 on
Domain Generalization
on NICO Vehicle
Adversarial Robustness for Unsupervised Domain Adaptation
Extensive Unsupervised Domain Adaptation (UDA) studies have shown great success in practice by learning transferable representations across a labeled source domain and an unlabeled target domain with deep models.
MultiSiam: Self-supervised Multi-instance Siamese Representation Learning for Autonomous Driving
By pre-training on SODA10M, a large-scale autonomous driving dataset, MultiSiam exceeds the ImageNet pre-trained MoCo-v2, demonstrating the potential of domain-specific pre-training.
Towards Understanding the Generative Capability of Adversarially Robust Classifiers
We find that adversarial training contributes to obtaining an energy function that is flat and has low energy around the real data, which is the key for generative capability.
G-DetKD: Towards General Distillation Framework for Object Detectors via Contrastive and Semantic-guided Feature Imitation
In this paper, we investigate the knowledge distillation (KD) strategy for object detection and propose an effective framework applicable to both homogeneous and heterogeneous student-teacher pairs.
NASOA: Towards Faster Task-oriented Online Fine-tuning with a Zoo of Models
Fine-tuning from pre-trained ImageNet models has been a simple, effective, and popular approach for various computer vision tasks.
AutoBERT-Zero: Evolving BERT Backbone from Scratch
Transformer-based pre-trained language models like BERT and its variants have recently achieved promising performance in various natural language processing (NLP) tasks.
Ranked #10 on
Semantic Textual Similarity
on MRPC
One Million Scenes for Autonomous Driving: ONCE Dataset
To facilitate future research on exploiting unlabeled data for 3D detection, we additionally provide a benchmark in which we reproduce and evaluate a variety of self-supervised and semi-supervised methods on the ONCE dataset.
SODA10M: A Large-Scale 2D Self/Semi-Supervised Object Detection Dataset for Autonomous Driving
Experiments show that SODA10M can serve as a promising pre-training dataset for different self-supervised learning methods, which gives superior performance when fine-tuning with different downstream tasks (i. e., detection, semantic/instance segmentation) in autonomous driving domain.
Adversarial Invariant Learning
To the best of our knowledge, this is one of the first to adopt differentiable environment splitting method to enable stable predictions across environments without environment index information, which achieves the state-of-the-art performance on datasets with strong spurious correlation, such as Colored MNIST.
Transformation Invariant Few-Shot Object Detection
To this end, we propose a simple yet effective Transformation Invariant Principle (TIP) that can be flexibly applied to various meta-learning models for boosting the detection performance on novel class objects.
Contextualizing Meta-Learning via Learning to Decompose
To this end, we propose Learning to Decompose Network (LeadNet) to contextualize the meta-learned ``support-to-target'' strategy, leveraging the context of instances with one or mixed latent attributes in a support set.
Towards a Theoretical Framework of Out-of-Distribution Generalization
We also introduce a new concept of expansion function, which characterizes to what extent the variance is amplified in the test domains over the training domains, and therefore give a quantitative meaning of invariant features.
Joint-DetNAS: Upgrade Your Detector with NAS, Pruning and Dynamic Distillation
For student morphism, weight inheritance strategy is adopted, allowing the student to flexibly update its architecture while fully utilize the predecessor's weights, which considerably accelerates the search; To facilitate dynamic distillation, an elastic teacher pool is trained via integrated progressive shrinking strategy, from which teacher detectors can be sampled without additional cost in subsequent searches.
TransNAS-Bench-101: Improving Transferability and Generalizability of Cross-Task Neural Architecture Search
While existing NAS methods mostly design architectures on a single task, algorithms that look beyond single-task search are surging to pursue a more efficient and universal solution across various tasks.
BWCP: Probabilistic Learning-to-Prune Channels for ConvNets via Batch Whitening
To address this problem, we develop a probability-based pruning algorithm, called batch whitening channel pruning (BWCP), which can stochastically discard unimportant channels by modeling the probability of a channel being activated.
How Well Does Self-Supervised Pre-Training Perform with Streaming Data?
Prior works on self-supervised pre-training focus on the joint training scenario, where massive unlabeled data are assumed to be given as input all at once, and only then is a learner trained.
iVPF: Numerical Invertible Volume Preserving Flow for Efficient Lossless Compression
We also propose a lossless compression algorithm based on iVPF.
Deeply Unsupervised Patch Re-Identification for Pre-training Object Detectors
Unsupervised pre-training aims at learning transferable features that are beneficial for downstream tasks.
SparseBERT: Rethinking the Importance Analysis in Self-attention
A surprising result is that diagonal elements in the attention map are the least important compared with other attention positions.
Loss Function Discovery for Object Detection via Convergence-Simulation Driven Search
For object detection, the well-established classification and regression loss functions have been carefully designed by considering diverse learning challenges.
DetCo: Unsupervised Contrastive Learning for Object Detection
Unlike most recent methods that focused on improving accuracy of image classification, we present a novel contrastive learning approach, named DetCo, which fully explores the contrasts between global image and local image patches to learn discriminative representations for object detection.
Out-of-Distribution Generalization Analysis via Influence Function
One of the definitions of OOD accuracy is worst-domain accuracy.