For most industries, like autonomous driving, security, and surveillance, infrastructure monitoring, etc. the data that needs to be annotated is collected via cameras. In the annotation stage, one would normally have to pre-process their data to fit the format of most image annotation platforms.
The new feature of video upload allows you to avoid the unnecessary process of splitting your footage into frames. The system does that work for you.
Most importantly, in the upload stage, you can also define the frame rate at which you want your videos to be uploaded. The system will then split your video into frames according to the specified frame rate.
The video can be upload in most common formats, such as:
The new feature makes the video upload as effortless as possible and is an important cornerstone of the video annotation process itself.
Image panel redesign
The new design for the image panel is specifically catered to accommodate both image and video annotation flow. Whether you have individual images or consecutive frames, the annotation process is much more intuitive and user friendly. The new image panel allows the user to:
- Have more editing space to focus on the image itself
- Transition smoothly through either consecutive or distant images
- Have the possibility to permanently unhide and hide the image tab throughout the project.
- Have a full display of image statuses
- A smoother and faster transition between images
- Have a full display of project progress by image status breakdown. Information like the total number of images in the project and the total number of images broken down by their status.
Following the Image tab redesign and the video upload option, the next improvement in the video annotation field is the Duplicate mode.
The new feature will allow you not only freely duplicate single or multiple objects for individual frames in any order but also will significantly cut the annotation time for video annotation projects, as it will allow you to continuously duplicate and edit recurring objects throughout multiple frames, without having to constantly switch between Duplicate and Edit mode.
The feature will allow you to add and subtract the objects you want to duplicate as you go, without disrupting the process and having to switch between modes.
The next advancement in the video annotation process is the interpolation mode. As a form of automation, the interpolation mode will make the object tracking in both video and image annotation projects faster and more efficient. Similar to Duplicate mode, the interpolation mode will allow you to track your objects in any frame range. With this option, the full manual annotation of consecutive frames can be replaced with the semi-automated process, where the annotator has to perform menial manual changes only.
We hope you enjoy the new SuperAnnotate updates. We are open to hear any feedback you have on them.