Joseph Turian's Projects
Implementation of neural language models, in particular Collobert + Weston (2008) and a stochastic margin-based version of Mnih's LBL.
Reference implementation of real-time autoregressive wavenet inference
OpenL3: Open-source deep audio and image embeddings
OpenL3 audio model weights, in numpy format
A hyperparameter optimization framework
OSQA branch, with some fixes
Install OSQA on webfaction
jsMath support for OSQA
A neural network with a sparse input, for predicting decisions of a natural language syntax parser.
Perceptual Metrics of Audio - perceptually relevant loss function
Pitch detection, for an audio file, using the Echonest remix API
How to solve the traveling salesman problem with the 2-opt algorithm, a fast heuristic search algorithm.
In Python, read the .80 file format, for 80legs web crawl results.
Python code for accessing the CrowdFlower API
Python methods to interact with the Crunchbase API v1.
Random projection library for Python, converting a dictionary to low-dimensional numpy matrix
Compress short strings, using the Huffman algorithm.
Example code for pytc (Python TokyoCabinet API)
Preprocess text for NLP (tokenizing, lowercasing, stemming, sentence splitting, etc.)
A permissive version of the Python SimpleXMLRPCServer, which can correct errant XML input from the client.
KDDCup 2005 query classification with word representations
KDDCup 2005 query classification with word representations
Induce word representations using random indexing (RI)
Instructions for using the RenderMan Python API for controlling the Dexed FM synthesizer on Linux
Extension for Mozilla Firefox and Google Chrome to save all of your open tabs to a text file (window/tab index, URL and title of each tab)
Simple Moog-like subtractive synth in SuperCollider, meant to be runnable from command line
scikit-learn main repo
Recipes for scikits.learn
Javascript autocomplete, with MySQL/PHP backend