31–40 of 42 posts
Unrelated to the software itself, but it is amazing how this website runs http without redirecting to https. You don't see that very often any more. It's probably an oversight, I guess :)
Earlier quoted context omitted.
When I last played with it I was able to get quite good results with an A6400 and a phone as the gyro (the camera does not natively provide the data), except for when the rolling shutter is extreme, at which point it falls apart. The camera's notoriously slow sensor readout seems to be more than Gyroflow can accommodate for when it gets pushed.
Why is gyro data required? Doesn't the video itself contain enough information about the movement in vast majority of cases?
It looks cool, though I still find myself looking for Hyperlapse[0] alternatives. Hyperlapse seemed like magic when announced, particularly, not needing to crop large parts of shaky videos [0] https://blogs.microsoft.com/ai/hyperlapse-siggraph-2014/
Johannes Kopf, the main author behind the original Hyperlapse paper, is one of the researchers behind Localrf discussed here yesterday [0]. I think Hyperlapse Pro [1], which stitched patches from multiple frames together to create new frames after doing some 3D reconstruction and then recreating a smoother camera path was mostly replaced with Hyperlapse mobile [2]. The Pro version was really slow, but the mobile vers…
This software is pure gold. I would actually love to give a donation to support its developement if possible.
It looks cool, though I still find myself looking for Hyperlapse[0] alternatives. Hyperlapse seemed like magic when announced, particularly, not needing to crop large parts of shaky videos [0] https://blogs.microsoft.com/ai/hyperlapse-siggraph-2014/
Uses Neural Radiance Fields
What do they mean by "gyro-based"?