Deep Audio Embedding
Visualizing the structure of Music in 3D
Recent posts
Predicting Fish Movement with Convolutional LSTMs part 5/5
Results and Discussion
Predicting Fish Movement with Convolutional LSTMs part 4/5
The model and training phase
Predicting Fish Movement with Convolutional LSTMs part 3/5
The experimental setup, dataset, and definition of input tensors and targets
Predicting Fish Movement with Convolutional LSTMs part 2/5
Outline of Project Subtasks
Predicting Fish Movement with Convolutional LSTMs part 1/5
Project motivation, guide to the github repository, and tools used
The Impact of Training data on Texture Synthesis models part 2/2
Using fully convolutional networks to synthesize textures from natural images
The Impact of Training data on Texture Synthesis models part 1/2
Using fully convolutional networks to synthesize textures from natural images
Putting zero-shot learning to the test with VAEs part 3/3: A Loss function for learning disentangled latent factors
Changing the VAE loss function to reduce latent variable covariance
Putting zero-shot learning to the test with VAEs part 2/3: Parameter Space Transects
Can a VAE reconstruct images from an unseen region of the latent space?
Putting zero-shot learning to the test with VAEs part 1/3: Building the dataset
Introduction and how I construct a known latent space training dataset
Parallelized Algorithms for computing large Similarity Matricies
A Comparison of the speed and memory use across three similarity matrix algorithms.
t-SNE Implementations with more flexible similarity metrics in the embedding space
An alternative embedding space similarity metric allows tunable emphasis on the data manifold’s global structure
Methods for graph creation from point clouds
K-nearest neighbors, Epsilon ball, Tau rule, and 2D lattice graphs
Computing the Concave Hull of a set of points in 2D
A Concave hull method built for generic 2D data
Collective object transport using multi-agent reinforcement learning
The impact of self-awareness on specialization in a cooperative multi-agent RL setting
Demonstrating Feasibility of Group-Level Rewards in a Toy Model
How rewards delivered ‘by proxy’ can teach agents to act in favor of group objectives