Well if you are shy, you have two options. Support going back to the office so everybody commutes there and you can easily talk to people without putting much effort... Or take it as a growth opportunity for yourself, take advantage of the extra time that you have saved in that useless commuting, and try to talk to people outside work and make friends as everybody else does.
This article could have been summarized in three paragraphs.
I'm really hating this trend of diluting content by giving useless testimonials, random anecdotes and delaying the resolution of the subject as much as possible.
This is the general idea of how I understood it works.
- It's based on WAN a video model, it generates a frame at a time in latent space that we later we decode into RGB
- We keep the latent space and its RGB of each frame in a database. Along with the RGB we compute depth so we get a cloud point (RGBD). This will be used for persistence.
- For each new frame we check which past frames have their point cloud contained in the camera frustum. We take the top 3 frames with more overlap and get their latent space.
- We feed these 3 frames to WAN via the cross attention layers for conditioning and that is how we achieve consistency
- The RGBD can be used to generate a gaussian splatting of the scene.
They updated it recently to fix the stuttering in the Disney app, and that issue had been there for a year or so. And they did that probably because Disney paid them.