awesome-multi-task-learning
A curated list of DATASETS, CODEBASES and PAPERS on Multi-Task Learning (MTL), from Machine Learning perspective.
https://github.com/thuml/awesome-multi-task-learning
Last synced: 28 days ago
JSON representation
Architecture
Decoder-focused Model
- TaskPrompter: Spatial-Channel Multi-Task Prompting for Dense Scene Understanding
- Inverted Pyramid Multi-task Transformer for Dense Scene Understanding
- Pattern-Affinitive Propagation Across Depth, Surface Normal and Semantic Segmentation
- PAD-Net: Multi-tasks Guided Prediction-and-Distillation Network for Simultaneous Depth Estimation and Scene Parsing
Hard Parameter Sharing
- MultiTask-CenterNet (MCN): Efficient and Diverse Multitask Learning using an Anchor Free Approach
- UberNet: Training a Universal Convolutional Neural Network for Low-, Mid-, and High-Level Vision Using Diverse Datasets and Limited Memory
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Layer by Layer: Uncovering Where Multi-Task Learning Happens in Instruction-Tuned Large Language Models
- Multitask Learning
- Multitask Learning
- Multitask Learning
- MultiNet: Real-time Joint Semantic Reasoning for Autonomous Driving
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
- Multitask Learning
Modularity, MoE, Routing & NAS
- Mod-Squad: Designing Mixture of Experts As Modular Multi-Task Learners
- M$^ 3$ViT: Mixture-of-Experts Vision Transformer for Efficient Multi-task Learning with Model-Accelerator Co-design
- An Evolutionary Approach to Dynamic Introduction of Tasks in Large-scale Multitask Learning Systems
- SkillNet-NLU: A Sparsely Activated Model for General-Purpose Natural Language Understanding
- Combining Modular Skills in Multitask Learning
- Multi-Task Reinforcement Learning with Soft Modularization
- MTL-NAS: Task-Agnostic Neural Architecture Search towards General-Purpose Multi-Task Learning
- Stochastic Filter Groups for Multi-Task CNNs: Learning Specialist and Generalist Convolution Kernels
- Deep Elastic Networks with Model Selection for Multi-Task Learning
- SNR: Sub-Network Routing for Flexible Parameter Sharing in Multi-Task Learning
- Flexible Multi-task Networks by Learning Parameter Allocation
- Feature Partitioning for Efficient Multi-Task Architectures
- Evolutionary architecture search for deep multitask networks
- NestedNet: Learning Nested Sparse Structures in Deep Neural Networks
- Learning Modular Neural Network Policies for Multi-Task and Multi-Robot Transfer
- MMoE
- PLE
- Factorizing Knowledge in Neural Networks
- AutoMTL: A Programming Framework for Automated Multi-Task Learning
- An Evolutionary Approach to Dynamic Introduction of Tasks in Large-scale Multitask Learning Systems
- Routing Networks: Adaptive Selection of Non-linear Functions for Multi-Task Learning
- Beyond Shared Hierarchies: Deep Multitask Learning through Soft Layer Ordering
- Modular Multitask Reinforcement Learning with Policy Sketches
- Learning Modular Neural Network Policies for Multi-Task and Multi-Robot Transfer
- PathNet: Evolution Channels Gradient Descent in Super Neural Networks
- SNR: Sub-Network Routing for Flexible Parameter Sharing in Multi-Task Learning
Modulation & Adapters
- Learning multiple visual domains with residual adapters
- Lossless Adaptation of Pretrained Vision Models For Robotic Manipulation
- Few-Shot Parameter-Efficient Fine-Tuning is Better and Cheaper than In-Context Learning
- Rethinking Hard-Parameter Sharing in Multi-Domain Learning
- The Power of Scale for Parameter-Efficient Prompt Tuning
- Prefix-Tuning: Optimizing Continuous Prompts for Generation
- Counter-Interference Adapter for Multilingual Machine Translation
- LoRA: Low-Rank Adaptation of Large Language Models
- AdapterFusion: Non-Destructive Task Composition for Transfer Learning
- A Study of Residual Adapters for Multi-Domain Neural Machine Translation
- AdapterHub: A Framework for Adapting Transformers
- MAD-X: An Adapter-Based Framework for Multi-Task Cross-Lingual Transfer
- Masking as an Efficient Alternative to Finetuning for Pretrained Language Models
- Many Task Learning With Task Routing
- BERT and PALs: Projected Attention Layers for Efficient Adaptation in Multi-Task Learning
- Efficient Parametrization of Multi-domain Deep Neural Networks
- A Study of Residual Adapters for Multi-Domain Neural Machine Translation
- MAD-X: An Adapter-Based Framework for Multi-Task Cross-Lingual Transfer
- Rethinking Hard-Parameter Sharing in Multi-Domain Learning
- Learning to Prompt for Continual Learning
- AdapterHub: A Framework for Adapting Transformers
Others
- Learning Sparse Sharing Architectures for Multiple Tasks
- Deep Asymmetric Multi-task Feature Learning
- Learning to Multitask
- Piggyback: Adapting a Single Network to Multiple Tasks by Learning to Mask Weights
- PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruning
- Asymmetric Multi-task Learning based on Task Relatedness and Confidence
- Deep Asymmetric Multi-task Feature Learning
- Asymmetric Multi-task Learning based on Task Relatedness and Confidence
Soft Parameter Sharing
- Latent Multi-task Architecture Learning
- NDDR-CNN: Layerwise Feature Fusing in Multi-Task CNNs by Neural Discriminative Dimensionality Reduction
- Learning Multiple Tasks with Multilinear Relationship Networks
- Cross-Stitch Networks for Multi-task Learning
- Progressive Neural Networks
- Deep Multi-task Representation Learning: A Tensor Factorisation Approach
- Trace Norm Regularised Deep Multi-Task Learning
- Trace Norm Regularised Deep Multi-Task Learning
Benchmark & Dataset
Computer Vision
- [URL
- Indoor Segmentation and Support Inference from RGBD Images
- [URL
- [URL
- [URL
- [URL
- [URL
- [URL
- [URL
- Omnidata: A Scalable Pipeline for Making Multi-Task Mid-Level Vision Datasets from 3D Scans
- [URL
- BDD100K: A Diverse Driving Dataset for Heterogeneous Multitask Learning
- MultiTask-CenterNet (MCN): Efficient and Diverse Multitask Learning using an Anchor Free Approach
- Omnidata: A Scalable Pipeline for Making Multi-Task Mid-Level Vision Datasets from 3D Scans