update SolidWorks assembly files

README.md

@@ -20,7 +20,7 @@ Some realworld experiment: [YouTube](https://youtu.be/LHvtbKmTwvE), [bilibili](h
 **Faster and Simpler:** The code is greatly simplified and refactored in Python/PyTorch. We also replaced the simulator with our CUDA-accelerated randomized environment, which is faster, lightweight, and boundless. For the stable version consistent with our paper, please refer to the [main](https://github.com/TJU-Aerial-Robotics/YOPO/tree/main) branch.
 
 ### Hardware:
-Our drone designed by [@Mioulo](https://github.com/Mioulo) is also open-source. The hardware components are listed in [hardware_list.pdf](hardware/hardware_list.pdf), and the SolidWorks file of carbon fiber frame can be found in [/hardware](hardware/).
+Our drone designed by [@Mioulo](https://github.com/Mioulo) is also open-source. The hardware components are listed in [hardware_list.pdf](hardware/hardware_list.pdf), and the SolidWorks file of carbon fiber frame can be found in [/hardware](hardware/) (complete assembly files are included in the [Release](https://github.com/TJU-Aerial-Robotics/YOPO/releases/tag/hardware)).
 
 ## Introduction:
 We propose **a learning-based planner for autonomous navigation in obstacle-dense environments** which integrates (i) perception and mapping, (ii) front-end path searching, and (iii) back-end optimization of classical methods into a single network. 
@@ -101,7 +101,7 @@ You can refer to [config.yaml](Simulator/src/config/config.yaml) for modificatio
 
 **3. Start the YOPO Planner** 
 
-You can refer to [traj_opt.yaml](YOPO/config/traj_opt.yaml) for modification of the flight speed (The given weights are pretrained at 6 m/s and perform smoothly at speeds between 0 - 6 m/s).
+You can refer to [traj_opt.yaml](YOPO/config/traj_opt.yaml) for modification of the flight speed (The given weights are pretrained at 6 m/s and perform smoothly at speeds between 0 - 6 m/s, and more pretrained models are available at [Releases](https://github.com/TJU-Aerial-Robotics/YOPO/releases)).
 
 ```
 cd YOPO

YOPO/loss/guidance_loss.py

@@ -66,7 +66,7 @@ class GuidanceLoss(nn.Module):
         goal_length = goal_dir.norm(dim=1)  # [B]
 
         # length difference along goal direction (cosine similarity)
-        parallel_diff = (goal_length - traj_along).abs()  # [B]
+        parallel_diff = F.smooth_l1_loss(goal_length, traj_along, reduction='none')
 
         # length perpendicular to goal direction
         traj_perp = traj_dir - traj_along.unsqueeze(1) * goal_dir_norm  # [B, 3]

YOPO/test_yopo_ros.py

@@ -140,7 +140,7 @@ class YopoNet:
 
         obs = np.concatenate((vel_c, acc_c, goal_c), axis=0).astype(np.float32)
         obs_norm = self.state_transform.normalize_obs(torch.from_numpy(obs[None, :]))
-        return obs_norm.to(self.device, non_blocking=True)
+        return obs_norm
 
     @torch.inference_mode()
     def callback_depth(self, data):
@@ -171,9 +171,8 @@ class YopoNet:
         # input prepare
         time1 = time.time()
         depth_input = torch.from_numpy(depth).to(self.device, non_blocking=True)  # (non_blocking: copying speed 3x)
-        obs_norm = self.process_odom()
+        obs_norm = self.process_odom().to(self.device, non_blocking=True)
         obs_input = self.state_transform.prepare_input(obs_norm)
-        obs_input = obs_input.to(self.device, non_blocking=True)
         # torch.cuda.synchronize()
 
         time2 = time.time()