Tags: vvrs/carla
Tags
CARLA 0.9.3
* Upgraded to Unreal Engine 4.21
* Upgraded Boost to 1.69.0
* New Town04 (biggest so far), includes a freeway, new bridge and road barrier, a nicer landscape based on height-map, and new street props
* New Town05, adding more variety of intersections for the scenario runner
* Redesigned pedestrian models and animations (walk and idle) for male and female characters
* Added sensor for detecting obstacles (ray-cast based)
* Added sensor GNSS (GPS)
* Basic agent integrated with global router
* Added a few methods to manage an actor:
- set_velocity: for setting the linear velocity
- set_angular_velocity: for setting the angular velocity
- get_angular_velocity: for getting the angular velocity
- add_impulse: for applying an impulse (in world axis)
* Renamed vehicle.get_vehicle_control() to vehicle.get_control() to be consistent with walkers
* Added new mesh for traffic lights
* Added new pine tree assets, with their LODs finely tuned for performance
* Added point transformation functionality for LibCarla and PythonAPI
* Added "sensor_tick" attribute to sensors (cameras and lidars) to specify the capture rate in seconds
* Added Export/Import map tools
* Added "get_forward_vector()" to rotation and transform, retrieves the unit vector on the rotation's X-axis
* Added support for Deepin in PythonAPI's setup.py
* Added support for spawning and controlling walkers (pedestrians)
* Updated BasicAgent to allow setting target_speed and handle US-style traffic lights properly
* OpenDriveActor has been rewritten using the Waypoint API, this has fixed some bugs
* Remove crash reporter from packaged build
* Improved simulator fatal error handling, now uses UE4 fatal error system
* LibCarla server pipeline now compiles with exceptions disabled for better performance and compatibility with UE4
* Fixed TCP accept error, too many open files while creating and destroying a lot of sensors
* Fixed lost error messages in client-side, now when a request fails it reports the reason
* Fixed global route planner to handle round about turns and made the code consistent with local planner
* Fixed local planner to avoid premature route pruning at path overlaps
* Fixed autopilot direction not properly initialized that interfered with the initial raycast direction
* Fixed crash when an actor was destroyed but not de-registered, e.g. falling out of world bounds
CARLA 0.9.2
* Updated ROS bridge for CARLA 0.9.X (moved to its own repository)
* Added Python API "agents" extension, includes
- Global route planner based on the Waypoints API (compatible with OpenDrive)
- BasicAgent: new client agent that can drive to a given coordinate of the map using the waypoint API and PID controllers, attending to other vehicles and traffic lights
- RoamingAgent: new client agent that can drive at different speeds following waypoints based on PID controllers, attending to other vehicles and traffic lights
- LocalPlanner functionality to navigate waypoints using PID controllers
- LateralControl and LongitudinalControl PIDs
* Add support for manual gear shifting
* Add "role_name" attribute to actors to easily identify the "hero" vehicle
* Updates to documentation and tutorials
* Simulator now starts by default in windowed mode
* CMake version required downgraded to 3.5 for better compatibility
* Fix waypoints height were all placed at zero height
* Fix actors in world.get_actors() missing parent actor
* Fix some vehicles losing their wheels after calling set_simulate_physics
* Several fixes and improvements to OpenDriveActor
CARLA 0.9.1
* New town: Town03
- Created with Vector Zero's RoadRunner (including OpenDrive information of the road layout)
- Bigger and more diverse
- More road variety: multiple lanes and lane markings, curves at different angles, roundabout, elevation, tunnel
* Lots of improvements to the Python API
- Support for Python 3
- Support for retrieving and changing lighting and weather conditions
- Migrated Lidar sensor
- Migrated image converter methods: Depth, LogarithmicDepth, and CityScapesPalette
- Migrated IO methods for sensor data, "save_to_disk" available for PNG, JPEG, TIFF, and PLY
- Added support for requesting the list of all the actors alive in the current world, `world.get_actors()`
- `world.get_actors()` returns an `ActorList` object with `filter` functionality and lazy initialization of actors
- Added collision event sensor, "sensor.other.collision", that triggers a callback on each collision to the actor it is attached to
- Added lane detector sensor, "sensor.other.lane_detector", that detects lane invasion events
- Added `carla.Map` and `carla.Waypoint` classes for querying info about the road layout
- Added methods for converting and saving the map as OpenDrive format
- Added `map.get_spawn_points()` to retrieve the recommended spawn points for vehicles
- Added `map.get_waypoint(location)` to query the nearest waypoint
- Added `map.generate_waypoints(distance)` to generate waypoints all over the map at an approximated distance
- Added `map.get_topology()` for getting a list the tuples of waypoints that define the edges of the road graph
- Added `waypoint.next(distance)` to retrieve the list of the waypoints at a distance that can be driven from this waypoint
- Added `parent` attributes to actors, not None if the actor is attached to another actor
- Added `semantic_tags` to actors containing the list of tags of all of its components
- Added methods for retrieving velocity and acceleration of actors
- Added function to enable/disable simulating physics on an actor, `actor.set_simulate_physics(enabled=True)`
- Added bounding boxes to vehicles, `vehicle.bounding_box` property
- Exposed last control applied to vehicles, `vehicle.get_vehicle_control()`
- Added a "tick" message containing info of all the actors in the scene
- Executed in the background and cached
- Added `world.wait_for_tick()` for blocking the current thread until a "tick" message is received
- Added `world.on_tick(callback)` for executing a callback asynchronously each time a "tick" message is received
- These methods return/pass a `carla.Timestamp` object containing, frame count, delta time of last tick, global simulation time, and OS timestamp
- Methods retrieving actor's info, e.g. `actor.get_transform()`, don't need to connect with the simulator, which makes these calls quite cheap
- Allow drawing debug shapes from Python: points, lines, arrows, boxes, and strings (`world.debug.draw_*`)
- Added id (id of current episode) and map name to `carla.World`
- Exposed traffic lights and signs as actors. Traffic lights have a specialized actor class that has the traffic light state (red, green, yellow) as property
- Added methods for accessing and modifying individual items in `carla.Image` (pixels) and `carla.LidarMeasurement` (locations)
- Added `carla.Vector3D` for (x, y, z) objects that are not a `carla.Location`
- Removed `client.ping()`, `client.get_server_version()` accomplishes the same
- Renamed `contains_X()` methods to `has_X()`
- Changed `client.set_timeout(seconds)` to use seconds (float) instead of milliseconds
- Allow iterating attributes of an Actor's Blueprint
- Fixed wildcard filtering issues, now "vehicle.*" or "*bmw*" patterns work too
- Fixed `actor.set_transform()` broken for attached actors
* More Python example scripts and improved the present ones
- Now all the scripts use the list of recommended spawn points for each map
- Renamed "example.py" to "tutorial.py", and updated it with latest changes in API
- Added timeout to the examples
- "manual_control.py" performance has been improved while having more measurements
- "manual_control.py" now has options to change camera type and position
- "manual_control.py" now has options to iterate weather presets
- "manual_control.py" now has a fancier HUD with lots of info, and F1 key binding to remove it
- Added "dynamic_weather.py" to change the weather in real-time (the one used in the video)
- Added "spawn_npc.py" to quickly add a lot of NPC vehicles to the simulator
- Added "spawn_npc.py --safe" to only add non-problematic vehicles
- "vehicle_gallery.py" also got some small fixes
* Asset and content improvements
- New vehicle: Lincoln MKZ 2017
- Refactored weather system, parametrized to make it easier to use
- Improved control of bikes and motorbikes, still not perfect but causes less accidents
- Added building block generator system
- Misc city assets: New building, tunnel columns, rail-road bridges, new textures, new urban props
- Adjusted vehicle physics and center of mass
- Adjusted the maximum distance culling for foliage
- Adjusted pedestrian animations and scale issues (not yet available with new API though)
- Improved map building blueprints, spline based asset repeaters, and wall building tools
- Replaced uses of Unreal's Foliage system with standard static meshes to work around a visual bug in Linux systems
- Fixed filenames too long when packing the project on Windows
- Fixed "SplineMeshRepeater" loses its collider mesh from time to time
- Standardized asset nomenclature
* New system for road information based on OpenDrive format
- Added new map classes for querying info about the road layout and topology
- Added methods for finding closest point on the road
- Added methods for generating and iterating waypoints based on the road layout
- Added OpenDrive parser to convert OpenDrive files to our map data structures
* Other miscellaneous improvements and fixes
- Fixed single channel Lidar crash (by @cwecht)
- Fixed command-line argument `-carla-settings` fails to load absolute paths (by @harlowja)
- Added an option to command-line to change quality level when launching the simulator, `-quality-level=Low`
- Added ROS bridge odometry message (by @ShepelIlya)
- New lens distortion shader, sadly not yet integrated with our cameras :(
- New Docker tutorial
- Disabled texture streaming to avoid issue of textures not loading in scene captures
- Adjusted scene capture camera gamma to 2.4
- Fixed leaking objects in simulation when despawning a vehicle. Now Pawn's controller is destroyed too if necessary when destroying an Actor
- Fixed overflow on platform time-stamp, now it uses `double`
- Upgraded @rpclib to fix crash when client exits too fast (rpclib/PR#167)
- Moved "PythonClient" inside deprecated folder to avoid confusion
- Refactored sensor related code
- New plugin system for sensors that simplifies adding sensors, mini-tutorial at carla-simulator#830
- Compile-time dispatcher for sensors and serializers
* Improvements to the streaming library
- Added multi-streams for streaming simultaneously to multiple clients (used by the "tick" message)
- Messages re-use allocated memory when possible
- Allows unsubscribing from a stream
- Fixed client receives interleaved sensor messages, some messages can be discarded if connection is too slow though
- Fixed streaming client fails to connect in Windows
- Fixed streaming client keeps trying to reconnect after destroying a sensor
* Refactored client C++ API
- Python GIL is released whenever possible to avoid blocking
- Fixed deadlock when closing the simulator while a client is connected
- Fixed crash on simulator shutdown if a client has connected at some point
- Set methods are now sent async which greatly improves performance in the client-side
- Vehicle control is cached and not sent if haven't changed
- Suppressed exceptions in destructors
* Other development improvements
- Improved Linux Makefile, fine-grained targets to reduce compilation times in development
- Workaround for "setup.py" to link against "libcarla_client.a" again (Linux only)
- Added support for ".gtest" file, each line of this file is passed to GTest executables as arguments when running `make check` targets
- Python eggs are also archived on Jenkins to easily get them without downloading the full package
- Added uncrustify config file for formatting UE4 C++ code
CARLA 0.9.0
* Upgraded to Unreal Engine 4.19
* Redesign of the networking architecture
- Allows any number of clients to connect simultaneously
- Now is possible to add and remove at any time any vehicle or camera
- Now is possible to control any vehicle or camera
- Now is possible to place cameras anywhere
- Reduced to two ports instead of three
- First port uses an RPC protocol based on [rpclib](http://rpclib.net/)
- Second port is for the streaming of the sensor data
* Redesign of the Python API
- Actors and sensors are now exposed in the API and can be independently controlled
- The Python module is built in C++, with significant performance gain in some operations
- Many functionality haven't been ported yet, so expect a lot of things missing
* Redesign of the build system to accommodate the changes in dependencies
- Everything can be done now with the Makefile
- For the moment only Linux is supported, sorry
* Massive clean up of all unused assets
* Some aesthetic fixes to the vehicles
CARLA 0.8.4 * Community contribution: ROS bridge by @laurent-george * New vehicle: Tesla Model 3 * Added an option to _"CarlaSettings.ini"_ to disable bikes and motorbikes * Fixed missing collision of vehicles introduced in 0.8.3 * Improved stability of bikes and motorbikes * Improved autopilot turning behaviour at intersections, now using front wheels positions as reference * Temporarily removed Kawasaki Ninja motorbikes because the model was having some stability issues
CARLA 0.8.3
* Added two-wheeled vehicles, 3 bicycles and 4 motorbikes
* Several art optimizations (CARLA is now about 10% faster)
- Improved the performance of vegetation assets, adjusted LOD and culling distance, set billboards where possible
- Drastically reduced the number of polygons of the landscape while keeping the original shape
- Removed some high-cost unnecessary assets
- Remodelled Mustang and NissanMicra, now with less polygons and materials, better textures and LOD
- Remodelled building SM_TerracedHouse_01, now with more polygons but less materials and better textures
* CARLA releases include now a Dockerfile for building docker images
* Change in HUD: replace "FPS" by "Simulation Step"
* The current map name is now included in the scene description message sent to the client
* Adapted "manual_control.py" and "view_start_positions.py" to use the map name sent by the simulator
* Improved the vehicle spawning algorithm, now it tries to spawn as much cars as possible even if there are not enough spawn points
* "Setup.sh" is now faster and accepts an argument to run multiple jobs in parallel
* Fixed foliage distance culling using wrong distance in "Low Mode"
* Fixed NissanMicra slightly turning left when driving straight
CARLA 0.8.2
* Revamped driving benchmark
- Changed name from benchmark to driving benchmark
- Fully Redesigned the architecture of the module
- Added a lot more documentation
- Now you can stop and resume the benchmarks you run
* Rolled back vehicle's location to the pivot of the mesh instead of the center of the bounding box
* Added relative transform of the vehicle's bounding box to the measurements, player and non-players
* Added "frame number" to each sensor measurement so it is possible to sync all the measurements based on the frame they are produced
* Improved vehicle spawner to better handle spawning failures
* Walkers use now a closer angle to detect vehicles, so they don't stop moving if a car passes nearby
* Fixed lighting artefact causing the road to change its brightness depending on the distance to the camera
* Fixed captured images overexposed in Low mode
* Fixed illegal character in asset name
* Fixed editing sun azimuth angle in CarlaWeadther.ini had no effect
* Fixed crash when using a non-standard image size in DirectX (Windows)
* Fixed issue with using multiple "SceneCaptureToDiskCamera"
CARLA 0.8.1 * New Python example for visualizing the player start positions * Fixed box extent of non-player agents was sent in centimeters instead of meters * Fixed speed limits were sent in km/h instead of m/s * Fixed issue in Volkswagen T2 wheels causing it to overturn
CARLA 0.8.0
* Upgraded to Unreal Engine 4.18
* Created our own pedestrian 3D models free to use and distribute
* Removed Epic's Automotive Materials dependencies
* 360 Lidars support (similar to Velodyne HDL-32E or VLP-16) thanks to Anton Pechenko (Yandex)
- Ray-cast based
- Configurable settings
- Added methods to save points to disk as PLY file
* Added quality level settings
- Low: low quality graphics, about 3 times faster with one camera
- Epic: best quality (as before)
* Measurements now use SI units
- Locations: m
- Speed: m/s
- Acceleration: m/s^2
- Collisions: kg*m/s
- Angles: degrees
* Added API methods to convert depth images to a point cloud
- New method "image_converter.depth_to_local_point_cloud"
- A supplementary image can be passed to attach colors to the points
- New client example generates a point cloud in world coordinates
- Added Transform class to Python API
* Performance optimizations
- Significant speed improvements in both Epic and Low modes
- Fixed materials and improved shaders for roads, architecture, sidewalks, foliage, landscapes, cars, walkers, reflections, water
- Execution of a set of Project and Engine parameters to improve performance (quality, vsync, AO, occlusion)
- Generation of the road pieces using static meshes and actors instead of a single actor with instanced meshes
- Improved performance since now is able to apply occlusion and draw distance
- Images are captured asynchronously in the render thread
- In asynchronous mode, images may arrive up to two frames later
- In synchronous mode, game thread is blocked until images are ready
- Blueprint code optimizations for vehicles, walkers, and splines
- Added a way to configure different quality levels with culling distance and materials configuration
* Refactored sensor related code to ease adding new sensors in the future
* Added vehicle box extent to player measurements
* Removed the player from the list of non-player agents
* Adjusted bounding boxes to vehicles' height
* Changed vehicles' center to match bounding box
* Added autopilot mode to manual_control.py
* Added quality level options to manual_control.py and client_example.py
* Replaced background landscape and trees by a matte painting
* Fixed road map generated some meshes twice
* Small improvements to Windows support
- Fixed issues with the Makefile
- Fixed asset names too long or containing special characters
CARLA 0.7.1
* New Python API module: Benchmark
- Defines a set of tasks and conditions to test a certain agent
- Contains a starting benchmark, CoRL2017
- Contains Agent Class: Interface for benchmarking AIs
* New Python API module: Basic Planner (Temporary Hack)
- Provide routes for the agent
- Contains AStar module to find the shortest route
* Other Python API improvements
- Converter class to convert between Unreal world and map units
- Metrics module to summarize benchmark results
* Send vehicle's roll, pitch, and yaw to client (orientation is now deprecated)
* New RoutePlanner class for assigning fixed routes to autopilot (IntersectionEntrance has been removed)
* Create a random engine for each vehicle, which greatly improves repeatability
* Add option to skip content download in Setup.sh
* Few small fixes to the city assets
PreviousNext