Sep 12, 2022
Stereolabs continues to push the boundaries of depth-sensing by introducing a ground-breaking new depth map with the new NEURAL DEPTH MODE. Based on multiple years of development, we are finally releasing this new depth mode to bring depth sensing applications to the next level.
The ZED SDK 3.7 also features a new lossless and hardware-based compression mode to record Lossless SVO files in real-time, new parameters for fine-tuning in the object detection module, and improvements for the other depth mode (Performance, Quality, and Ultra).
Finally, we are also introducing performance improvements for the SVO recording and Streaming module.
Starting from 3.7.1, the Jetson installers will use the Linux4Tegra version instead of the JetPack version used previously for clarity.
From 3.7.2, adding beta support for JetPack 5.0 (L4T 34.1). From 3.7.7 adding stable support for JetPack 5.0 (L4T 35.1).
From 3.7.5, adding support for Ubuntu 22.04 and CUDA 11.7, the AI module accuracy and performance have been improved.
For older releases and changelog, see the ZED SDK release archive.
Maintenance mode versions (legacy)
These versions are too old and no longer fully supported, the AI module is not available for these platforms.
grab() triggers an internal recovery. The camera settings were set back to default, whereas they should have been kept as the previous chosen value. This is now the expected behavior.grab() function was triggering an internal recovery and lost control of camera settings.sl::Mat GPU allocation on Jetson leading to NO_GPU_COMPATIBLE errorPERSON_HEAD_BOX models inference speed by up to 50% while slightly improving accuracy, especially in the far rangeMULTI_CLASS models accuracy by ~2% while also improving inference speed, by up to 5%udev rules, to open the IMU without root access, were incorrect and could allow privilege escalation.get_data() function by removing the deep copy by default (up to 30% faster samples)Mat classenableRecording() / enableStreaming() in a separated thread than grab() function.NEURAL depth mode if necessary when requested.grab() recovery for more stability.NEURAL depth mode model.sl::Camera::getSensorsData() that could cause a segmentation fault.sl::BODY_FORMAT::POSE_34. This was causing unexpected jumps in body tracking.sl::Camera methods.NEURAL depth mode based on AI.sl::InitParameters::depth_mode = sl::DEPTH_MODE::NEURAL.sl::SVO_COMPRESSION_MODE::H264_LOSSLESS and sl::SVO_COMPRESSION_MODE::H265_LOSSLESS.sl::RuntimeParameters::remove_saturated_area to activate/de-activate saturation filtering during depth sensing.Introduced new object detection parameter with sl::ObjectDetectionParameters::filtering_mode associated with the new enumerate sl:OBJECT_FILTERING_MODE.true to keep the same behavior as the previous version, where saturated areas were removed from the depth map. You can now set it to false to keep those areas in the depth processing.sl:OBJECT_FILTERING_MODE::NMS3D to keep the same behavior as the previous version. You can now also choose NONE or NMS3D_PER_CLASS to adjust the level of filtering done during object tracking.ERROR_CODE Camera::getCurrentMinMaxDepth(float& min,float& max)grab() time.--cut option that created wrong compression artifacts on specific positions.set_on_parameters_set_callback with add_on_set_parameters_callback to improve compatibility with ROS 2 Galactic