Currently PhD student at Berkeley. Previously EECS at MIT BS '20 & MEng '21.
Highlights
- Pro
Block or Report
Block or report ethanweber
Report abuse
Contact GitHub support about this user’s behavior. Learn more about reporting abuse.
Report abusePopular repositories
-
IncidentsDataset Public
Code and data (Incidents Dataset) for ECCV 2020 Paper "Detecting natural disasters, damage, and incidents in the wild".
-
scaling-anno Public
Code and data for "Scaling up instance annotation via label propagation" in ICCV 2021.
-
ros-multimaster-app Public
This is an application that utilizes roslibjs websockets to synchronize ROS topics and services across multiple ROS masters on the same network.
JavaScript 3
556 contributions in the last year
Less
More
Contribution activity
October 2022
Created 41 commits in 1 repository
Created a pull request in nerfstudio-project/nerfstudio that received 14 comments
Opened 38 other pull requests in 1 repository
nerfstudio-project/nerfstudio
35
merged
1
open
2
closed
- Add script filenames to CLI docs
- Improving checkpoint logic
- Improve process data with filename filters and saving mp4s
- Google Colab support with viewer
- process data to grab only JPS and PNG, disregard the small trash files
- bump version
- save renders to a folder and bump viewer version
- update camera opt defaults
- fix optimizer dict
- v0.1.2
- turn on camera opt by default
- integrating record3d and fix checkpoint delete issue
- Pose optimization
- semantic nerfw docs update
- max number of iterations logic
- update method commands
- remove print
- doubling nerfacto samples on eval
- update image for viewer docs
- docs update
- add desolation scene
- improving the pipeline docs
- remove creating new methods docs
- change default websocket port
- save when paused viewer works now
- Some pull requests not shown.
Reviewed 24 pull requests in 1 repository
nerfstudio-project/nerfstudio
24 pull requests
- Reduce memory usage of nerfacto model
- Terrance/camera controls
- Reducing flicker artifacts for nerfacto method
- Link fix in README
- process data to grab only JPS and PNG, disregard the small trash files
- Added up vector offset
- fixed --start-train flag
- adding keyboard camera control instructions to viewer
- add video link to record3d
- Update custom_dataset.md
- v0.1.1
- Don't apply near plane during eval
- Check ffmpeg before rendering video
- Fix typos in README
- Remove outdated viewer logging
- Update colmap install guide
- Readme Changes
- Pushing more introduction docs
- Installation Docs Updates
- Record3D Bugfix: Removing Pose OpenGL Conversion
- Data download script fix
- Cameras pr followup
- Quickstart Guide Fixes
- Cameras
Created an issue in nerfstudio-project/nerfstudio that received 7 comments
Opened 10 other issues in 1 repository
nerfstudio-project/nerfstudio
4
open
6
closed
- Ability to save more than one checkpoint.
- "Too many users have viewed or downloaded this file recently."
- Server and render show disconnected when running training locally
- camera controls should only in effect when the mouse is over the three.js viewport
- add button to hide just the training images
-
change
?websocket_url=localhost:xxxxto?websocket_port=xxxx -
could we do e.g.,
ns-train --model nerfacto --dataparser record3d --help? - SceneBounds -> SceneBox naming
- merge MLP(FieldModule) with TCNN by putting in the same class
-
don't download to
/tmpindownload_data.py




