Effecient 2D SLAM lidar packages

edit

I'm working in a project that requires SLAM algorithms. I'm targeting 2D SLAM LiDAR for feature-based maps. My experience was with extended Kalman filter (EKF)-SLAM but I think this approach is outdated. What I really need two things from SLAM. First, I need the estimates of the robot's pose with its uncertainties. Second, the estimates of the obstacles' locations and their uncertainties. EKF provides me with this information but I need efficient SLAM package for higher accuracy and object detection for extracting the features. Do you suggest any package for this regard? I'm using ROS melodic version and Gazebo 9 in Ubuntu 18.04.