Differentiable Convex

Differentiable convex programming integrates convex optimization problems directly into neural networks, enabling end-to-end training of systems requiring both prediction and optimization. Current research focuses on developing efficient algorithms, such as Lagrangian Proximal Gradient Descent and ADMM-based methods, to handle the differentiable optimization layers within various architectures, including those employing Control Barrier Functions and Quadratic Programs. This approach is proving valuable in diverse applications like robotics, control systems, and resource allocation, offering improved performance and safety guarantees compared to traditional two-stage approaches that decouple prediction and optimization. The ability to seamlessly integrate optimization into learning systems is significantly advancing the capabilities of AI in safety-critical domains.

Papers

October 8, 2024

Differentiation Through Black-Box Quadratic Programming Solvers
Connor W. Magoon, Fengyu Yang, Noam Aigerman, Shahar Z. Kovalsky
Differentiable Neural Implicit Differentiation Auto Differentiation Quadratic Programming Differentiable Convex Mathematical Programming

July 22, 2024

Differentiable Convex Polyhedra Optimization from Multi-view Images
Daxuan Ren, Haiyi Mei, Hezi Shi, Jianmin Zheng, Jianfei Cai, Lei Yang
Multi View Image Mesh Reconstruction Shape Representation Differentiable Convex

July 8, 2024

LPGD: A General Framework for Backpropagation through Embedded Optimization Layers
Anselm Paulus, Georg Martius, Vít Musil
Gradient Descent Back Propagation General Framework Proximal Gradient Descent Optimization Layer Differentiable Convex Parametric Optimization

June 20, 2024

Differentiable-Optimization Based Neural Policy for Occlusion-Aware Target Tracking
Houman Masnavi, Arun Kumar Singh, Farrokh Janabi-Sharifi
Policy Optimization Conditional Variational Autoencoder Differentiable Convex Collision Avoidance Capability

January 30, 2024

LeTO: Learning Constrained Visuomotor Policy with Differentiable Trajectory Optimization
Zhengtong Xu, Yu She
Trajectory Optimization Problem Visuomotor Policy State of the Art Imitation Differentiable Convex Differentiable Trajectory

December 15, 2023

Constrained Meta-Reinforcement Learning for Adaptable Safety Guarantee with Differentiable Convex Programming
Minjae Cho, Chuangchuang Sun
Meta Reinforcement Learning Safety Constraint Safety Guarantee Differentiable Convex Convex Q Learning Learning Enabled

October 4, 2023

Soft Convex Quantization: Revisiting Vector Quantization with Convex Optimization
Tanmay Gautam, Reid Pryzant, Ziyi Yang, Chenguang Zhu, Somayeh Sojoudi
Convex Optimization Vector Quantization Vector Quantized Variational Autoencoder Differentiable Convex

October 3, 2023

Distributionally Safe Reinforcement Learning under Model Uncertainty: A Single-Level Approach by Differentiable Convex Programming
Alaa Eddine Chriat, Chuangchuang Sun
Reinforcement Learning Model Uncertainty Safety Constraint Single Stage Differentiable Convex

October 2, 2023

Decision-Oriented Learning Using Differentiable Submodular Maximization for Multi-Robot Coordination
Guangyao Shi, Chak Lam Shek, Nare Karapetyan, Pratap Tokekar
LeArning Abstract Submodular Maximization Greedy Algorithm Multi Robot Coordination Differentiable Convex Decision Aware Multi Robot Decision Making

September 15, 2023

Wasserstein Distributionally Robust Control Barrier Function using Conditional Value-at-Risk with Differentiable Convex Programming
Alaa Eddine Chriat, Chuangchuang Sun
Control Barrier Function Safety Constraint Safety Guarantee Distributionally Robust Value at Risk Differentiable Convex

May 31, 2023

Power Control with QoS Guarantees: A Differentiable Projection-based Unsupervised Learning Framework
Mehrazin Alizadeh, Hina Tabassum
Deep Neural Network Resource Allocation Quality of Service Power Control Differentiable Convex Interference Channel Differentiable Projection

April 12, 2023

Learning Robust and Correct Controllers from Signal Temporal Logic Specifications Using BarrierNet
Wenliang Liu, Wei Xiao, Calin Belta
Control Barrier Function Robust Learning Neural Network Controller Signal Temporal Logic Differentiable Convex Meta Controller

March 29, 2023

Meta-Learning Parameterized First-Order Optimizers using Differentiable Convex Optimization
Tanmay Gautam, Samuel Pfrommer, Somayeh Sojoudi
Meta Learning Better Optimizers Meta Learning Framework Differentiable Convex Step Optimization

November 25, 2022

End-to-End Stochastic Optimization with Energy-Based Model
Lingkai Kong, Jiaming Cui, Yuchen Zhuang, Rui Feng, B. Aditya Prakash, Chao Zhang
Stochastic Optimization Energy Based Model Stochastic Optimization Problem Optimization Landscape Differentiable Convex

July 1, 2022

DiffPills: Differentiable Collision Detection for Capsules and Padded Polygons
Kevin Tracy, Taylor A. Howell, Zachary Manchester
Trajectory Optimization Gradient Based Collision Detection 3D Capsule Differentiable Convex Differentiable Collision Harmonic Polygon

June 1, 2022

Learning to Sequence and Blend Robot Skills via Differentiable Optimization
Noémie Jaquier, You Zhou, Julia Starke, Tamim Asfour
LeArning Abstract Sequence of Sequence Robot Skill Differentiable Optimization Differentiable Convex Modular Skill Planar Robot Atomic Skill

December 14, 2021

Efficient differentiable quadratic programming layers: an ADMM approach
Andrew Butler, Roy Kwon
Optimization Algorithm Alternating Direction Method of Multiplier Differentiable Neural Quadratic Program Differentiable Convex

November 22, 2021

BarrierNet: A Safety-Guaranteed Layer for Neural Networks
Wei Xiao, Ramin Hasani, Xiao Li, Daniela Rus
Neural Network Control Barrier Function Neural Network Controller Differentiable Convex Safety Layer

Differentiable Convex

Papers

Differentiation Through Black-Box Quadratic Programming Solvers

Differentiable Convex Polyhedra Optimization from Multi-view Images

LPGD: A General Framework for Backpropagation through Embedded Optimization Layers

Differentiable-Optimization Based Neural Policy for Occlusion-Aware Target Tracking

LeTO: Learning Constrained Visuomotor Policy with Differentiable Trajectory Optimization

Constrained Meta-Reinforcement Learning for Adaptable Safety Guarantee with Differentiable Convex Programming

Soft Convex Quantization: Revisiting Vector Quantization with Convex Optimization

Distributionally Safe Reinforcement Learning under Model Uncertainty: A Single-Level Approach by Differentiable Convex Programming

Decision-Oriented Learning Using Differentiable Submodular Maximization for Multi-Robot Coordination

Wasserstein Distributionally Robust Control Barrier Function using Conditional Value-at-Risk with Differentiable Convex Programming

Power Control with QoS Guarantees: A Differentiable Projection-based Unsupervised Learning Framework

Learning Robust and Correct Controllers from Signal Temporal Logic Specifications Using BarrierNet

Meta-Learning Parameterized First-Order Optimizers using Differentiable Convex Optimization

End-to-End Stochastic Optimization with Energy-Based Model

DiffPills: Differentiable Collision Detection for Capsules and Padded Polygons

Learning to Sequence and Blend Robot Skills via Differentiable Optimization

Efficient differentiable quadratic programming layers: an ADMM approach

BarrierNet: A Safety-Guaranteed Layer for Neural Networks