Mitsubishi Electric Research Laboratories (MERL)'s repositories
cocktail-fork-separation
Baseline multi-resolution cross network model trained using the Divide and Remaster Dataset
hyper-unmix
Source code for training models and using the hyperbolic interface proposed in our ICASSP 2023 paper, “Hyperbolic Audio Source Separation.”
neural-IIR-field
Neural IIR Filter Field for HRTF Upsampling and Personalization
Sound2Sight
Generating Visual Dynamics from Sound and Context for generating video frames and their motion dynamics conditioned on audio and a few past frames
DeepBornFNO
Code for training and using the Born Fourier Neural Operator (BornFNO) model proposed in our ICASSP 2023 paper
CASENet
C++/Python implementation of our CASENet based on Caffe for multi-label semantic edge detection training/testing
CAZSL
Context-Aware Zero Shot Learning
CISOR
Convergent Inverse Scattering using Optimization and Regularization for a systematic comparison between CISOR, SEAGLE, and other state-of-the-art methods on simulated data
CME
Circular Maze Environment
FoldingNet_Plus
For unsupervised learning of 3D point clouds with data-adaptive graph prior
PCQM
Point Cloud Quality Metric for measuring the geometry distortion of point cloud compression
QNTRPO
Quasi-Newton Trust Region Policy Optimization
robust-rotation-estimation
Robust frame-to-frame camera rotation estimation in crowded scenes
SafetyRL
Goal directed RL with Safety Constraints