Network Security Internet Technology Development Database Servers Mobile Phone Android Software Apple Software Computer Software News IT Information

In addition to Weibo, there is also WeChat

Please pay attention

WeChat public account

Shulou

Tesla pushes FSD Beta firmware to North American vehicles equipped with HW4 hardware

2025-04-13 Update From: SLTechnology News&Howtos shulou NAV: SLTechnology News&Howtos > IT Information >

Share

Shulou(Shulou.com)11/24 Report--

CTOnews.com July 5 news, Tesla pushed FSD Beta 2023.20.7 (11.3.6) firmware to North American vehicles equipped with HW4 hardware, but this function button is still not open. However, considering Musk's statement, this feature should be available to North American users in the near future.

CTOnews.com with Tesla FSD Beta v11.3.6 (2023.20.7) Release Log:

- FSD Beta is enabled on the highway. This unified vision and planning systems and replaced the highway system that was more than four years old. While the old highway system still relied on a few single-camera and single-frame networks to handle simple lane-related operations, FSD Beta uses multi-camera video networks and next-generation processors that allow the system to handle more complex vehicle interaction behaviors, reducing lane dependence, and making way for smarter behavior, smoother control, and better decision-making.

- By mining an additional 30,000 auto-labeled segments from the test fleet, the recall rate for close cuts was improved by 15%, especially for large trucks and high yaw rate scenarios. In addition, dedicated speed control for cut objects has been extended and adjusted.

- Improved vehicle positioning in wide lanes, allowing other vehicles to steer around themselves by shifting their position in the direction of the upcoming turn.

- By shifting in lanes, handling in scenes with high curvature or large trucks is improved to maintain a safe distance from other vehicles on the road and improve comfort.

- Improved behavior in the case of lane changes in path congestion in congested traffic. Ego will now maintain a greater safety distance on blocked lanes to prevent possible congestion.

- Lane changes are improved in congested traffic scenarios, allowing vehicles to more naturally select gaps and overtake on adjacent lanes next to Ego by allowing higher acceleration during alignment phases.

- Smoother turns by improving detection consistency between lane, line and road edge predictions. This is achieved by integrating the latest version of the lane guidance module into the road edge and line network.

- improve that detection accuracy of the motion semantics of other vehicle. Accuracy improved by 23 percent when other vehicles transitioned to driving, and error decreased by 12 percent when Autopilot incorrectly identified the vehicle in front as parked. These improvements were achieved by increasing the video context in the network, adding more data for these scenes, and adding loss penalties to control the associated vehicles.

- Expands the maximum trajectory optimization field of view for smoother control on highly curved roads and at high speeds for distant vehicles.

- Adjacent to a row of parked vehicles in a narrow lane improves driving behavior, tending to drift within the lane rather than unnecessarily change lanes or slow down.

- Improved performance of continuous lane change maneuvers through better integration of vision-based positioning and coarse map lane counting.

- Text prompts have been added to the user interface to inform FSD Beta of upcoming manipulation actions. It also improves the visualization of impending deceleration in the vehicle path. Arrows appear in different degrees of transparency and speed to indicate the intensity of the deceleration, while solid lines indicate where the vehicle will stop.

- The recall and accuracy of object detection were improved, especially by reducing the position error of semi-trucks by 10%, improving the recall and accuracy of crossing vehicles with a distance of more than 100 meters by 3% and 7%, respectively, and improving the recall rate of motorcycles by 5%. This was achieved by implementing additional quality checks in our 2 million video clip auto-annotation dataset.

- By improving object motion modeling in low-speed scenes, false offsets near wide lanes and intersections are reduced.

- Adjusted the position of the automatic blind spot camera when FSD Beta is active to prioritize Autopilot visualizations. You can drag the camera head to save custom positions.

The FSD beta program started in October 2020 and has been updated since, with the launch of FSD Beta V11, Tesla's FSD has been adapted for urban and highway driving. So far, V11 has rolled out several updates.

A few days ago, Elon Musk said that FSD V12 will no longer be a "beta" when it is released, which may mean that Tesla's fully autonomous driving (FSD) software seems to be ending its "beta" phase and opening commercial.

Related Reading:

Musk: Tesla will offer FSD free trials for all cars in North America in the future, and then promote them to other regions after doing well locally.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.

Views: 0

*The comments in the above article only represent the author's personal views and do not represent the views and positions of this website. If you have more insights, please feel free to contribute and share.

Share To

IT Information

Wechat

© 2024 shulou.com SLNews company. All rights reserved.

12
Report