About
Helping companies automate data mapping using AI. Backed by General Catalyst, Khosla…
Experience
Education
Courses
-
AP Calculus BC
-
-
AP Chemistry
-
-
AP Computer Science A
-
-
AP Computer Science Principles
-
-
AP Macroeconomics
-
-
AP Physics 1
-
-
AP Physics C: Mechanics
-
-
AP Spanish Language and Culture
-
-
AP Statistics
-
-
Applied Matrix Methods
CME103
-
Computer Organization and Systems
CS107
-
Deep Learning
CS230
-
Linear Algebra
-
-
Mathematical Foundations of Computing
CS103
-
Multivariable Calculus
-
-
Practical Unix
CS1U
-
Programming Abstractions
CS106B
-
Standard C++ Programming
CS106L
-
The Wireless World, and the Data You Leak
EE26N
Projects
-
A Recurrent Neural Network for Neurons: Continuous Decoding of Intracortical Brain Signals for BMI Applications
Brain Machine Interfaces (BMI) are becoming feasible clinical treatments for paralyzed patients. Using only thought, patients are able to move robotic arms with increasing precision and dexterity. This project builds on existing linear methods to continuously decode a 192-dimensional binary neural signals into 2-dimensional X-Y positional coordinates.
A quick summary of the network architecture. We utilized a series of LSTM cells, where each cell took an input feature vector of 192…Brain Machine Interfaces (BMI) are becoming feasible clinical treatments for paralyzed patients. Using only thought, patients are able to move robotic arms with increasing precision and dexterity. This project builds on existing linear methods to continuously decode a 192-dimensional binary neural signals into 2-dimensional X-Y positional coordinates.
A quick summary of the network architecture. We utilized a series of LSTM cells, where each cell took an input feature vector of 192 elements. The output of each cell was then passed through 2 fully connected ReLU layers to reduce the output to a size 2 element vector.Other creatorsSee project -
Parsing Lecture Videos by Theme
Utilizing Python, this project timestamps and links each specific theme to an appropriate section within the video.
Using Houndify audio to text API, it converts the audio of the lecture video to a JSON object while marking the time by inputing markers to signify every few seconds. Then using Natural Language Processing, it segments the the video into themes. After the JSON is parsed, it uses the markers within each string to estimate the location within the video. Therefore allowing…Utilizing Python, this project timestamps and links each specific theme to an appropriate section within the video.
Using Houndify audio to text API, it converts the audio of the lecture video to a JSON object while marking the time by inputing markers to signify every few seconds. Then using Natural Language Processing, it segments the the video into themes. After the JSON is parsed, it uses the markers within each string to estimate the location within the video. Therefore allowing students to easily find specific lecture material.Other creators
Languages
-
Spanish
Limited working proficiency
-
English
Native or bilingual proficiency
Other similar profiles
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore MoreOthers named Robert Ross in United States
2902 others named Robert Ross in United States are on LinkedIn
See others named Robert Ross