Liangchen Li
Machine Learning and Machine Intelligence
University of Cambridge
Biography
This website is currently only used as a personal recorder of paper reading.
Interests
- Deep Learning Theory
- Computer Vision
- Robust Learning
- Anything else EXCEPT Natural Language Processing
Posts
The Campfire Tales
Please see original post for full version. Neural Tangent Kernel (NTK) is well known as a powerful tool that elegantly proves the convergence and generalisation of neural networks. However, as an engineering student who basically know nothing about mathematics, I find the original NTK paper a little bit opaque to understand. Driven by my intrinsic inquisitiveness (“vegetable but addicted”), I spent some time trying to figure out what on earth is NTK with every effort and managed to get an impression on that at last. This article aims to help people who are in the same boat, vegetable, but still struggling for a free ride in deep learning theory.
Readings
Full List (339)
Only includes thoroughly read papers. Full collection is much larger than this.
Foundation Models
Lightweight Large Language Models (5)
- Octo-planner: On-device Language Model for Planner-Action Agents (arXiv 2406)
- Octopus v4: Graph of language models (arXiv 2404)
- Octopus v3: Technical Report for On-device Sub-billion Multimodal AI Agent (arXiv 2404)
- Octopus v2: On-device language model for super agent (arXiv 2404)
- Octopus: On-device language model for function calling of software APIs (arXiv 2404)
Usage of Foundation Models (2)
Vision-Language Pretraining and Multimodal Large Language Models (38)
- LLaVA-ϕ: Efficient Multi-Modal Assistant with Small Language Model (arXiv 2401)
- MobileVLM : A Fast, Strong and Open Vision Language Assistant for Mobile Devices (arXiv 2312)
- InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks (arXiv 2312, CVPR 2024)
- Vary: Scaling up the Vision Vocabulary for Large Vision-Language Models (arXiv 2312)
- LLaVA-Interactive: An All-in-One Demo for Image Chat, Segmentation, Generation and Editing (arXiv 2311)
- MiniGPT-v2: large language model as a unified interface for vision-language multi-task learning (arXiv 2310, ICLR 2024 submission)
- PaLI-3 Vision Language Models: Smaller, Faster, Stronger (arXiv 2310, ICLR 2024 submission)
- Improved Baselines with Visual Instruction Tuning (arXiv 2310)
- InternLM-XComposer: A Vision-Language Large Model for Advanced Text-image Comprehension and Composition (arXiv 2309)
- Kosmos-2.5: A Multimodal Literate Model (arXiv 2309)
- Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond (arXiv 2308)
- Shikra: Unleashing Multimodal LLM’s Referential Dialogue Magic (arXiv 2306)
- Kosmos-2: Grounding Multimodal Large Language Models to the World (arXiv 2306)
- Improving CLIP Training with Language Rewrites (arXiv 2305, NeurIPS 2023)
- PaLI-X: On Scaling up a Multilingual Vision and Language Model (arXiv 2305)
- VisionLLM: Large Language Model is also an Open-Ended Decoder for Vision-Centric Tasks (arXiv 2305)
- InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning (arXiv 2305, NeurIPS 2023)
- LLaMA-Adapter V2: Parameter-Efficient Visual Instruction Model (arXiv 2304)
- MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models (arXiv 2304)
- Visual Instruction Tuning (arXiv 2304, NeurIPS 2023)
- LLaMA-Adapter: Efficient Fine-tuning of Language Models with Zero-init Attention (arXiv 2303)
- Language Is Not All You Need: Aligning Perception with Language Models (arXiv 2302)
- BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models (arXiv 2301)
- Scaling Language-Image Pre-training via Masking (arXiv 2212, CVPR 2023)
- Uni-Perceiver v2: A Generalist Model for Large-Scale Vision and Vision-Language Tasks (arXiv 2211, CVPR 2023)
- PaLI: A Jointly-Scaled Multilingual Language-Image Model (arXiv 2209, ICLR 2023)
- Flamingo: a Visual Language Model for Few-Shot Learning (arXiv 2204, NeurIPS 2022)
- BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation (arXiv 2201, ICML 2022)
- SLIP: Self-supervision meets Language-Image Pre-training (arXiv 2112, ECCV 2022)
- Uni-Perceiver: Pre-training Unified Architecture for Generic Perception for Zero-shot and Few-shot Tasks (arXiv 2112, CVPR 2022)
- LiT: Zero-Shot Transfer with Locked-image Text Tuning (arXiv 2111, CVPR 2022)
- FILIP: Fine-grained Interactive Language-Image Pre-Training (arXiv 2111, ICLR 2022)
- An Empirical Study of Training End-to-End Vision-and-Language Transformers (arXiv 2111, CVPR 2022)
- Align before Fuse: Vision and Language Representation Learning with Momentum Distillation (arXiv 2107, NeurIPS 2021)
- Multimodal Few-Shot Learning with Frozen Language Models (arXiv 2106, NeurIPS 2021)
- LoRA: Low-Rank Adaptation of Large Language Models (arXiv 2106, ICLR 2022)
- Learning Transferable Visual Models From Natural Language Supervision (arXiv 2103, ICML 2021)
- Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision (arXiv 2102, ICML 2021)
Image Generation
Autoregressive Models for Image Generation (4)
- Autoregressive Image Generation without Vector Quantization (arXiv 2406)
- An Image is Worth 32 Tokens for Reconstruction and Generation (arXiv 2406)
- Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction (arXiv 2404)
- Taming Transformers for High-Resolution Image Synthesis (CVPR 2021, arXiv 2012)
Consistency Models (8)
- Bidirectional Consistency Models (arXiv 2403)
- Multistep Consistency Models (arXiv 2403)
- SCott: Accelerating Diffusion Models with Stochastic Consistency Distillation (arXiv 2403)
- Trajectory Consistency Distillation: Improved Latent Consistency Distillation by Semi-Linear Consistency Function with Trajectory Mapping (arXiv 2402, PLAGIARIZED from the CTM paper)
- Improved Techniques for Training Consistency Models (arXiv 2310, ICLR 2024)
- Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inference (arXiv 2310, ICLR 2024 rejection)
- Consistency Trajectory Models: Learning Probability Flow ODE Trajectory of Diffusion (arXiv 2310, ICLR 2024)
- Consistency Models (arXiv 2303, ICML 2023)
Diffusion Models (30)
- PaGoDA: Progressive Growing of a One-Step Generator from a Low-Resolution Diffusion Teacher (arXiv 2405)
- Applying Guidance in a Limited Interval Improves Sample and Distribution Quality in Diffusion Models (arXiv 2404)
- Fine-grained Text-to-Image Synthesis with Semantic Refinement (ICLR 2024 rejection, IJCAI 2024 submission)
- Self-conditioned Image Generation via Generating Representations (arXiv 2312)
- Analyzing and Improving the Training Dynamics of Diffusion Models (arXiv 2312, CVPR 2024)
- Adversarial Diffusion Distillation (arXiv 2311)
- Aligning Text-to-Image Diffusion Models with Reward Backpropagation (arXiv 2310, ICLR 2024 rejection)
- SDXL: Improving Latent Diffusion Models for High-Resolution Image Synthesis (arXiv 2307, ICLR 2024)
- Your Diffusion Model is Secretly a Zero-Shot Classifier (arXiv 2303, ICML 2023 workshop, ICCV 2023)
- Understanding Diffusion Objectives as the ELBO with Simple Data Augmentation (arXiv 2303, NeurIPS 2023)
- I2SB: Image-to-Image Schrödinger Bridge (arXiv 2302, ICML 2023)
- Adding Conditional Control to Text-to-Image Diffusion Models (arXiv 2302, CVPR 2023)
- Inversion-Based Style Transfer with Diffusion Models (arXiv 2211, CVPR 2023)
- eDiff-I: Text-to-Image Diffusion Models with an Ensemble of Expert Denoisers (arXiv 2211)
- Classifier-Free Diffusion Guidance (arXiv 2207, NeurIPS 2021 workshop)
- Elucidating the Design Space of Diffusion-Based Generative Models (arXiv 2206, NeurIPS 2022)
- Dual Diffusion Implicit Bridges for Image-to-Image Translation (arXiv 2203, ICLR 2023)
- MaskGIT: Masked Generative Image Transformer (arXiv 2202, CVPR 2022)
- Progressive Distillation for Fast Sampling of Diffusion Models (arXiv 2202, ICLR 2022)
- High-Resolution Image Synthesis with Latent Diffusion Models (arXiv 2112, CVPR 2022)
- Palette: Image-to-Image Diffusion Models (arXiv 2111, SIGGRAPH 2022)
- SDEdit: Guided Image Synthesis and Editing with Stochastic Differential Equations (arXiv 2108, ICLR 2022)
- Structured Denoising Diffusion Models in Discrete State-Spaces (arXiv 2107, NeurIPS 2021)
- Diffusion Models Beat GANs on Image Synthesis (arXiv 2105, NeurIPS 2021)
- Image Super-Resolution via Iterative Refinement (arXiv 2104, TPAMI 2022)
- Score-Based Generative Modeling through Stochastic Differential Equations (arXiv 2011, ICLR 2021)
- Denoising Diffusion Implicit Models (arXiv 2006, ICLR 2021)
- Denoising Diffusion Probabilistic Models (arXiv 2006, NeurIPS 2020)
- Improved Techniques for Training Score-Based Generative Models (arXiv 2006, NeurIPS 2020)
- Generative Modeling by Estimating Gradients of the Data Distribution (arXiv 1907, NeuIPS 2019)
Generative Adversarial Networks (17)
- Idempotent Generative Network (arXiv 2311, ICLR 2024 submission)
- Alias-Free Generative Adversarial Networks (arXiv 2106, NeurIPS 2021)
- Analyzing and Improving the Image Quality of StyleGAN (arXiv 1912, CVPR 2020)
- A Style-Based Generator Architecture for Generative Adversarial Networks (arXiv 1812, CVPR 2019)
- Large Scale GAN Training for High Fidelity Natural Image Synthesis (arXiv 1809, ICLR 2019)
- Self-Attention Generative Adversarial Networks (arXiv 1805, ICML 2019)
- Spectral Normalization for Generative Adversarial Networks (arXiv 1802, ICLR 2018)
- StackGAN++: Realistic Image Synthesis with Stacked Generative Adversarial Networks (arXiv 1710, TPAMI 2018)
- Progressive Growing of GANs for Improved Quality, Stability, and Variation (arXiv 1710, ICLR 2018)
- Improved Training of Wasserstein GANs (arXiv 1704, NeurIPS 2017)
- Wasserstein Generative Adversarial Networks (arXiv 1701, ICML 2017)
- Towards Principled Methods for Training Generative Adversarial Networks (arXiv 1701, ICLR 2017)
- StackGAN: Text to Photo-realistic Image Synthesis with Stacked Generative Adversarial Networks (arXiv 1612, ECCV 2017)
- Improved Techniques for Training GANs (arXiv 1606, NeurIPS 2016)
- Unsupervised Representation Learning with Deep Convolutional Generative Adversarial Networks (arXiv 1511, ICLR 2016)
- Conditional Generative Adversarial Nets (arXiv 1411)
- Generative Adversarial Nets (arXiv 1406, NeurIPS 2014)
Vision Tasks
Image Segmentation (2)
Intrinsic Image Decomposition, Re-lighting, Harmonization and Shadow Removal (19)
- Relighting Humans in the Wild: Monocular Full-Body Human Relighting with Domain Adaptation (arXiv 2110, PG 2021)
- SSH: A Self-Supervised Framework for Image Harmonization (arXiv 2108, ICCV 2021)
- CANet: A Context-Aware Network for Shadow Removal (arXiv 2108, ICCV 2021)
- Unsupervised Portrait Shadow Removal via Generative Priors (arXiv 2108, ACM MM 2021)
- Single-image Full-body Human Relighting (arXiv 2107, EGSR 2021)
- Region-aware Adaptive Instance Normalization for Image Harmonization (arXiv 2106, CVPR 2021)
- GMLight: Lighting Estimation via Geometric Distribution Approximation (arXiv 2102, AAAI 2021)
- Physics-based Shading Reconstruction for Intrinsic Image Decomposition (arXiv 2009, CVIU 2021)
- Total Relighting: Learning to Relight Portraits for Background Replacement (TOG 2021)
- Neural Light Transport for Relighting and View Synthesis (arXiv 2008, TOG 2021)
- Learning to Factorize and Relight a City (arXiv 2008, ECCV 2020)
- PIFuHD: Multi-Level Pixel-Aligned Implicit Function for High-Resolution 3D Human Digitization (arXiv 2004, CVPR 2020)
- Unsupervised Learning of Probably Symmetric Deformable 3D Objects from Images in the Wild (arXiv 1911, CVPR 2020, TPAMI 2021)
- DoveNet: Deep Image Harmonization via Domain Verification (arXiv 1911, CVPR 2020)
- Unsupervised Learning for Intrinsic Image Decomposition from a Single Image (arXiv 1911, CVPR 2020)
- Relighting Humans: Occlusion-Aware Inverse Rendering for Full-Body Human Images (arXiv 1908, SIGGRAPH ASIA 2018)
- SfSNet: Learning Shape, Reflectance and Illuminance of Faces in the Wild (arXiv 1712, CVPR 2018)
- 3D Object Manipulation in a Single Photograph using Stock 3D Models (SIGGRAPH 2014)
- Intrinsic Image Decomposition Using a Sparse Representation of Reflectance (TPAMI 2013)
Virtual Try-on, pose estimation, 3D reconstruction and Graphics Related (16)
- Shape Controllable Virtual Try-on for Underwear Models (arXiv 2107)
- Learning High Fidelity Depths of Dressed Humans by Watching Social Media Dance Videos (arXiv 2103, CVPR 2021)
- Disentangled Cycle Consistency for Highly-realistic Virtual Try-On (arXiv 2103, CVPR 2021)
- TryOnGAN: Body-Aware Try-On via Layered Interpolation (arXiv 2101, SIGGRAPH 2021)
- Learning to Transfer Texture from Clothing Images to 3D Humans (arXiv 2003, CVPR 2020)
- Warping Residual Based Image Stitching for Large Parallax (CVPR 2020)
- Towards Photo-Realistic Virtual Try-On by Adaptively Generating-Preserving Image Content (CVPR 2020)
- Generating High-Resolution Fashion Model Images Wearing Custom Outfits (arXiv 1908, ICCV 2019 Workshop on Computer Vision for Fashion, Art and Design)
- Multi-Garment Net: Learning to Dress 3D People from Images (arXiv 1908, ICCV 2019)
- Expressive Body Capture: 3D Hands, Face, and Body From a Single Image (CVPR 2019)
- DensePose: Dense Human Pose Estimation In The Wild (arXiv 1802, CVPR 2018)
- A Simple yet Effective Baseline for 3d Human Pose Estimation (ICCV 2017, arXiv 1705)
- Stacked Hourglass Networks for Human Pose Estimation (ECCV 2016, arXiv 1603)
- SMPL: A Skinned Multi-Person Linear Model (TOG 2015)
- Shape deformation using skeleton correspondences for realistic posed fashion flat creation (IEEE Transactions on Automation Science and Engineering 2014)
- Image Deformation Using Moving Least Squares (SIGGRAPH 2006)
- A Theory of Shape by Space Carving (IJCV 2000)
GAN Prior and GAN Inversion (21)
- Feature-Style Encoder for Style-Based GAN Inversion (arXiv 2202, CVPR 2022 submission)
- High-Fidelity GAN Inversion for Image Attribute Editing (arXiv 2109)
- Designing an encoder for StyleGAN image manipulation (TOG 2021)
- ReStyle: A Residual-Based StyleGAN Encoder via Iterative Refinement (arXiv 2104, ICCV 2021)
- A Simple Baseline for StyleGAN Inversion (arXiv 2104)
- Surrogate Gradient Field for Latent Space Manipulation (arXiv 2104, CVPR 2021)
- GAN Inversion: A Survey (arXiv 2101)
- Improved StyleGAN Embedding: Where are the Good Latents? (arXiv 2012)
- Improving the Fairness of Deep Generative Models without Retraining (arXiv 2012)
- Encoding in style: a stylegan encoder for image-to-image translation (arXiv 2008, CVPR 2021)
- Generative Hierarchical Features from Synthesizing Images (arXiv 2007, CVPR 2021)
- Closed-Form Factorization of Latent Semantics in GANs (arXiv 2007, CVPR 2021)
- InterFaceGAN: Interpreting the Disentangled Face Representation Learned by GANs (arXiv 2005, TPAMI 2020)
- In-Domain GAN Inversion for Real Image Editing (arXiv 2004, ECCV 2020)
- Editing in Style: Uncovering the Local Semantics of GANs (arXiv 2004, CVPR 2020)
- Exploiting Deep Generative Prior for Versatile Image Restoration and Manipulation (arXiv 2003, ECCV 2020, TPAMI 2021)
- Image Processing Using Multi-Code GAN Prior (arXiv 1912, CVPR 2020)
- Semantic Hierarchy Emerges in Deep Generative Representations for Scene Synthesis (arXiv 1911, IJCV 2020)
- Disentangled Inference for GANs with Latently Invertible Autoencoder (arXiv 1906)
- Image2StyleGAN: How to Embed Images Into the StyleGAN Latent Space? (arXiv 1904, ICCV 2019)
- Deep Image Prior (arXiv 1711, CVPR 2018)
Image Translation and Processing (9)
- Blind Face Restoration via Deep Multi-scale Component Dictionaries (arXiv 2008, ECCV 2020)
- STGAN: A Unified Selective Transfer Network for Arbitrary Image Attribute Editing (arXiv 1904, CVPR 2019)
- AttGAN: Facial Attribute Editing by Only Changing What You Want (arXiv 1711, TIP 2019)
- StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation (arXiv 1711, CVPR 2018)
- High-Resolution Image Synthesis and Semantic Manipulation with Conditional GANs (arXiv 1711, CVPR 2018)
- Arbitrary Style Transfer in Real-time with Adaptive Instance Normalization (arXiv 1703, ICCV 2017)
- Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks (arXiv 1703, ICCV 2017)
- Image-to-Image Translation with Conditional Adversarial Networks (arXiv 1611, CVPR 2017)
- U-Net: Convolutional Networks for Biomedical Image Segmentation (arXiv 1505, MICCAI 2015)
Vision Backbones (7)
- An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale (arXiv 2010, ICLR 2021)
- Densely Connected Convolutional Networks (arXiv 1608, CVPR 2017)
- Wide Residual Networks (arXiv 1605, BMVC 2016)
- Deep Residual Learning for Image Recognition (arXiv 1512, CVPR 2016)
- Very Deep Convolutional Networks for Large-Scale Image Recognition (arXiv 1409, ICLR 2015)
- Going Deeper with Convolutions (arXiv 1409, CVPR 2015)
- Imagenet classification with deep convolutional neural networks (NeurIPS 2012)
Robust Learning
Data Poisoning (5)
- Robust Unlearnable Examples: Protecting Data Privacy Against Adversarial Learning (ICLR 2022)
- Fooling Adversarial Training with Inducing Noise (arXiv 2111, ICLR 2022 rejection)
- Adversarial examples make strong poisons (arXiv 2106, NeurIPS 2021)
- Unlearnable Examples: Making Personal Data Unexploitable (arXiv 2101, ICLR 2021)
- Witches' Brew: Industrial Scale Data Poisoning via Gradient Matching (arXiv 2009, ICLR 2021)
Adversarial Examples and Attacks (32)
- Learning to Learn Transferable Attack (arXiv 2112, AAAI 2022)
- Mind the box: l1-APGD for sparse adversarial attacks on image classifiers (arXiv 2103, ICML 2021)
- Enhancing the Transferability of Adversarial Attacks through Variance Tuning (arXiv 2103, CVPR 2021)
- Backpropagating Linearly Improves Transferability of Adversarial Examples (arXiv 2012, NeurIPS 2020)
- Practical No-box Adversarial Attacks against DNNs (arXiv 2012, NeurIPS 2020)
- Reliable evaluation of adversarial robustness with an ensemble of diverse parameter-free attacks (arXiv 2003, ICML 2020)
- Skip Connections Matter: On the Transferability of Adversarial Examples Generated with ResNets (arXiv 2002, ICLR 2020)
- An Alternative Surrogate Loss for PGD-based Adversarial Testing (arXiv 1910)
- Adversarial Examples are Just Bugs, Too (distill 1908)
- Minimally distorted Adversarial Examples with a Fast Adaptive Boundary Attack (arXiv 1907, ICML 2020)
- Enhancing Adversarial Example Transferability with an Intermediate Level Attack (arXiv 1907, ICCV 2019)
- Adversarial Examples Are Not Bugs, They Are Features (arXiv 1905, NeurIPS 2019)
- Evading Defenses to Transferable Adversarial Examples by Translation-invariant Attacks (arXiv 1904, CVPR 2019)
- WITCHcraft: Efficient PGD attacks with random step size (arXiv 1904, ICASSP 2020)
- Disentangling Adversarial Robustness and Generalization (arXiv 1812, CVPR 2019)
- Are adversarial examples inevitable? (arXiv 1809, ICLR 2019)
- Constructing Unrestricted Adversarial Examples with Generative Models (arXiv 1805, NeurIPS 2018)
- Improving Transferability of Adversarial Examples with Input Diversity (arXiv 1803, CVPR 2019)
- Adversarial Risk and the Dangers of Evaluating Against Weak Attacks (arXiv 1802, ICML 2018)
- Obfuscated Gradients Give a False Sense of Security: Circumventing Defenses to Adversarial Examples (arXiv 1802, ICML 2018)
- Boosting adversarial attacks with momentum (arXiv 1710, CVPR 2018)
- Towards Deep Learning Models Resistant to Adversarial Attacks (arXiv 1706, ICLR 2018)
- The Space of Transferable Adversarial Examples (arXiv 1704)
- Delving into Transferable Adversarial Examples and Black-box Attacks (arXiv 1611, ICLR 2017)
- Universal adversarial perturbations (arXiv 1610, CVPR 2017)
- Towards Evaluating the Robustness of Neural Networks (arXiv 1608, IEEE SP 2017)
- Adversarial examples in the physical world (arXiv 1607, ICLR 2017 workshop)
- Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samples (arXiv 1605)
- DeepFool: a simple and accurate method to fool deep neural networks (arXiv 1511, CVPR 2016)
- Explaining and Harnessing Adversarial Examples (arXiv 1412, ICLR 2015)
- Intriguing properties of neural networks (arXiv 1312, ICLR 2014)
- Evasion Attacks against Machine Learning at Test Time (arXiv 1708, ECCV 2013)
Robust Learning and Adversarial Defenses (56)
- Balance, Imbalance, and Rebalance: Understanding Robust Overfitting from a Minimax Game Perspective (arXiv 2310, NeurIPS 2023)
- Improving Generalization of Adversarial Training via Robust Critical Fine-Tuning (arXiv 2308, ICCV 2023)
- Revisiting Outer Optimization in Adversarial Training (ECCV 2022, arXiv 2209)
- Can we achieve robustness from data alone? (arXiv 2207, ICML 2022 workshop, ICLR 2023 withdrawl)
- Catastrophic Overfitting Can Be Induced with Discriminative Non-robust Features (arXiv 2206, TMLR 2023)
- Understanding Robust Overfitting of Adversarial Training and Beyond (arXiv 2206, ICML 2022)
- Diffusion Models for Adversarial Purification (arXiv 2205, ICML 2022)
- Transcoders: A Better Alternative To Denoising Autoencoders (CVPR 2022 submission)
- An Eye for an Eye: Defending against Gradient-based Attacks with Gradients (arXiv 2202, CVPR 2022 submission)
- Learning Sample Reweighting for Adversarial Robustness (ICLR 2022 rejection)
- Self-ensemble Adversarial Training for Improved Robustness (ICLR 2022)
- Data Augmentation Can Improve Robustness (arXiv 2111, NeurIPS 2021)
- Exploring Memorization in Adversarial Training (arXiv 2106, ICLR 2022)
- Clustering Effect of Adversarial Robust Models (NeurIPS 2021)
- Relating Adversarially Robust Generalization to Flat Minima (arXiv 2104, ICCV 2021)
- Consistency Regularization for Adversarial Robustness (arXiv 2103, ICML workshop AML 2021, AAAI 2022)
- ZeroGrad : Mitigating and Explaining Catastrophic Overfitting in FGSM Adversarial Training (arXiv 2103)
- Improving Adversarial Robustness via Channel-wise Activation Suppressing (arXiv 2103, ICLR 2021)
- Robust Overfitting May Be Mitigated by Properly Learned Smoothening (ICLR 2021)
- Recent Advances in Adversarial Training for Adversarial Robustness (arXiv 2102, IJCAL 2021)
- Data Quality Matters For Adversarial Training: An Empirical Study (arXiv 2101, ICLR 2022 rejection)
- Advocating for Multiple Defense Strategies against Adversarial Examples (arXiv 2012, ECML PKDD 2020 workshops)
- Are Adversarial Examples Created Equal? A Learnable Weighted Minimax Risk for Robustness under Non-uniform Attacks (arXiv 2010, AAAI 2021)
- Do Wider Neural Networks Really Help Adversarial Robustness? (arXiv 2010, NeurIPS 2021)
- Geometry-aware Instance-reweighted Adversarial Training (arXiv 2010, ICLR 2021)
- Bag of Tricks for Adversarial Training (arXiv 2010, ICLR 2021)
- Adversarial Feature Desensitization (arXiv 2006, NeurIPS 2021)
- Adversarial Weight Perturbation Helps Robust Generalization (arXiv 2004, NeurIPS 2020)
- Manifold Regularization for Locally Stable Deep Neural Networks (arXiv 2003)
- Attacks Which Do Not Kill Training Make Adversarial Learning Stronger (arXiv 2002, ICML 2020)
- Overfitting in adversarially robust deep learning (arXiv 2002, ICML 2020)
- Fast is better than free: Revisiting adversarial training (arXiv 2001, ICLR 2020)
- Improving Adversarial Robustness Requires Revisiting Misclassified Examples (ICLR 2020)
- Adversarially Robust Representations with Smooth Encoders (ICLR 2020)
- API-Net: Robust Generative Classifier via a Single Discriminator (ECCV 2020)
- Adversarial Distributional Training for Robust Deep Learning (NeurIPS 2020)
- Adversarial Examples Improve Image Recognition (arXiv 1911, CVPR 2020)
- Invert and Defend: Model-based Approximate Inversion of Generative Adversarial Networks for Secure Inference (arXiv 1911)
- Adversarial Robustness through Local Linearization (arXiv 1907, NeurIPS 2019)
- Intriguing properties of adversarial training at scale (arXiv 1906, ICLR 2020)
- Interpreting Adversarially Trained Convolutional Neural Networks (arXiv 1905, ICML 2019)
- Adversarial Training for Free! (arXiv 1904, NeurIPS 2019)
- Adversarial Training and Robustness for Multiple Perturbations (arXiv 1904, NeurIPS 2019)
- Improving Adversarial Robustness via Promoting Ensemble Diversity (arXiv 1901, ICML 2019)
- Theoretically Principled Trade-off between Robustness and Accuracy (arXiv 1901, ICML 2019)
- MMA Training: Direct Input Space Margin Maximization through Adversarial Training (arXiv 1812, ICLR 2020)
- Feature Denoising for Improving Adversarial Robustness (arXiv 1812, CVPR 2019)
- Robustness via curvature regularization, and vice versa (arXiv 1811, CVPR 2019)
- Evaluating and Understanding the Robustness of Adversarial Logit Pairing (arXiv 1807, NeurIPS SECML 2018)
- Robustness May Be at Odds with Accuracy (arXiv 1805, ICLR 2019)
- Adversarial Logit Pairing (arXiv 1803)
- Gradient Regularization Improves Accuracy of Discriminative Models (arXiv 1712, Schedae Informaticae 2018)
- Ensemble Adversarial Training: Attacks and Defenses (arXiv 1705, ICLR 2018)
- MagNet: a Two-Pronged Defense against Adversarial Examples (arXiv 1705, ACM SIGSAC CCS 2017)
- On detecting adversarial perturbations (arXiv 1702, ICLR 2017)
- Adversarial Machine Learning at Scale (arXiv 1611, ICLR 2017)
Others
Autonomous Driving (2)
AI for Biomedical Science (29)
- DeSAM: Decoupling Segment Anything Model for Generalizable Medical Image Segmentation (arXiv 2306)
- How to Efficiently Adapt Large Segmentation Model(SAM) to Medical Images (arXiv 2306)
- The Impact of ChatGPT and LLMs on Medical Imaging Stakeholders: Perspectives and Use Cases (arXiv 2306)
- LLaVA-Med: Training a Large Language-and-Vision Assistant for Biomedicine in One Day (arXiv 2306)
- PMC-VQA: Visual Instruction Tuning for Medical Visual Question Answering (arXiv 2305)
- Echo from noise: synthetic ultrasound image generation using diffusion models for real image segmentation (arXiv 2305, ASMUS 2023)
- Towards Segment Anything Model (SAM) for Medical Image Segmentation: A Survey (arXiv 2305)
- BiomedGPT: A Unified and Generalist Biomedical Generative Pre-trained Transformer for Vision, Language, and Multimodal Tasks (arXiv 2305)
- UniverSeg: Universal Medical Image Segmentation (arXiv 2304, ICCV 2023)
- Customized Segment Anything Model for Medical Image Segmentation (arXiv 2304)
- Medical SAM Adapter: Adapting Segment Anything Model for Medical Image Segmentation (arXiv 2304)
- Segment Anything Model for Medical Images? (arXiv 2304)
- Segment Anything in Medical Images (arXiv 2304)
- MI-SegNet: Mutual Information-Based US Segmentation for Unseen Domain Generalization (arXiv 2303)
- Large-Scale Domain-Specific Pretraining for Biomedical Vision-Language Processing (arXiv 2303)
- PMC-CLIP: Contrastive Language-Image Pre-training using Biomedical Documents (arXiv 2303, MICCAI 2023)
- RAMM: Retrieval-augmented Biomedical Visual Question Answering with Multi-modal Pre-training (arXiv 2303)
- Towards Unifying Medical Vision-and-Language Pre-training via Soft Prompts (arXiv 2302)
- Unifying Molecular and Textual Representations via Multi-task Language Modelling (arXiv 2301, ICML 2023)
- Medical Visual Question Answering via Conditional Reasoning and Contrastive Learning (TMI 2023)
- Self-Supervised Vision-Language Pretraining for Medical Visual Question Answering (arXiv 2211, ISBI 2023)
- BioGPT: Generative Pre-trained Transformer for Biomedical Text Generation and Mining (arXiv 2210, Briefings in Bioinformatics 2022)
- Multi-Modal Masked Autoencoders for Medical Vision-and-Language Pre-Training (arXiv 2209, MICCAI 2022)
- Diffusion Deformable Model for 4D Temporal Medical Image Generation (arXiv 2206, MICCAI 2022)
- Does CLIP Benefit Visual Question Answering in the Medical Domain as Much as it Does in the General Domain? (arXiv 2112)
- GLoRIA: A Multimodal Global-Local Representation Learning Framework for Label-Efficient Medical Image Recognition (ICCV 2021)
- Cooperative Training and Latent Space Data Augmentation for Robust Medical Image Segmentation (arXiv 2107, MICCAI 2021)
- Medical Visual Question Answering via Conditional Reasoning (ACM MM 2020)
- Overcoming Data Limitation in Medical Visual Question Answering (arXiv 1909, MICCAI 2019)
Explainable AI (1)
Robotics (2)
Knowledge Distillation (2)
Data Augmentation (2)
Visualisation (1)
Convergence and Generalization (5)
- Convergence of Adversarial Training in Overparametrized Neural Networks (arXiv 1906, NeurIPS 2019)
- On the Convergence and Robustness of Adversarial Training (ICML 2019, arXiv 2112)
- A Convergence Theory for Deep Learning via Over-Parameterization (arXiv 1811, ICML 2019)
- Neural Tangent Kernel: Convergence and Generalization in Neural Networks (arXiv 1806, NeurIPS 2018)
- Deep Neural Networks as Gaussian Processes (arXiv 1711, ICLR 2018)
Topological Data Analysis (7)
- Topological Autoencoders (arXiv 1906, ICML 2020)
- Characterizing the Shape of Activation Space in Deep Neural Networks (arXiv 1901, ICMLA 2019)
- Neural Persistence: A Complexity Measure for Deep Neural Networks Using Algebraic Topology (arXiv 1812, ICLR 2019)
- Topological Data Analysis of Decision Boundaries with Application to Model Selection (arXiv 1805, ICML 2019)
- On Characterizing the Capacity of Neural Networks using Algebraic Topology (arXiv 1802)
- Adversary Detection in Neural Networks via Persistent Homology (arXiv 1711)
- Persistence stability for geometric complexes (arXiv 1207, Geometriae Dedicata 2014)
Contrastive/Self-supervised/Unsupervised/Representation Learning (5)
- Supervised Contrastive Learning (arXiv 2004, NeurIPS 2020)
- A Simple Framework for Contrastive Learning of Visual Representations (arXiv 2002, ICML 2020)
- Learning Disentangled Representations via Mutual Information Estimation (arXiv 1912, ECCV 2020)
- Data-Efficient Image Recognition with Contrastive Predictive Coding (arXiv 1905, ICML 2020)
- Representation Learning with Contrastive Predictive Coding (arXiv 1807)
Organization and Architecture (2)
Deep Reinforcement Learning (3)
- Modular Architecture for StarCraft II with Deep Reinforcement Learning (arXiv 1811, AAAI 2018)
- Implementation of Q Learning and Deep Q Network For Controlling a Self Balancing Robot Model (arXiv 1807, Robotics and Biomimetics 2018)
- Playing Atari with Deep Reinforcement Learning (arXiv 1312, NeuralIPS Deep Learning Workshop 2013)
