Hola a tod@s
Recibí hoy las respuestas a las preguntas que hicieron, por parte de Kartaview:
" Q: What is the vision for the functionality to build a 360-degree panoramic photo, from taking several pictures in the same point? What is the plan for the iOS version?
A: We are not planning to have this functionality in the near future, but we are supporting uploading 360 images captured by other apps/cameras via scripts.
Q: Is/will it possible to edit the location of the photos once uploaded? A functionality planned to fix manually the outlayer images in sequences.
A: We do not currently support this capability, but we are looking at adding this functionality.
Q: Is it planned to make an upload platform for images taken from sports cameras, user-friendly and multiplatform?
A: We are supporting this via scripts (GitHub - kartaview/upload-scripts: Uploader tools for KartaView). We have not planned for GUI yet, but if there is interest from the community side to improve this, we are willing to support it.
Q: GitHub suggests that traffic sign detection is no longer available in the U.S. (Traffic sign detection is not yet available for this track · Issue #259 · kartaview/openstreetcam.org · GitHub) What is the status of these features across countries?
We are looking to re-starting detectors for the US, but can not promise due to budget constraints. Cloud cost is the main limiting factor.
Q: It hasn’t been communicated at all for a while - what is the medium-term plan to collaborate with communities?
We can allocate support toward community initiatives if needed.
Q: Is there a plan like in the past for Grant Cameras?
There isn’t a plan in place for this aspect at the moment, however, depending on the magnitude of the project you’re planning on running, we can discuss certain options.
Q: What is the medium and long-term plan for maintenance?
We will update our repository with the most recent features, as well as track, respond to and fix the reported issues on GitHub. On kartaview.org we will keep improving our training sets for Traffic signs, Personally identifiable info, and POIs.
Q: What is the current and future use of the photos by KartaView (train algorithms that recognize objects)?
KartaView images serve as a ground-truthing mechanism to improve OpenStreetMap by editing new features (e.g. lanes, turn restrictions, road name) in the map or validating the existing ones, especially while doing armchair mapping. Once recordings reach our server, they will go through a complex procedure that involves snapping photos to the roads, removing Personal Identifiable Information (faces and license plates), and preparing them to be published to kartaview.org. After publishing images on kartaview.org users can edit OSM from KartaView imagery and detections. They can also tag traffic signs and check automatic detections to improve KartaView’s training sets. KartaView runs three types of object detections on uploaded images: Traffic signs, Personally identifiable info, and POIs.
Q: What part of the code is open? Where is the (updated) code made available?
iOS, mobile application code is open. To be more precise, the iOS repo has the latest features but overall it is missing our latest releases that were focused on bug fixing.
The Android repo is behind in terms of features, so we’ll have some work to do before we can post updates to this repository. Aside from the code, we try to have the most recent APKs available for download from our repo. The upload scripts repo is now open; we’ve noticed a lot of interest in this area over the years, and we aim to be as responsive as possible given our engineering capabilities."