Skip to content

Fix X-NavDP clutter evaluation scaling #116

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
yty-sky wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Fix X-NavDP clutter evaluation scaling #116

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion baselines/x-navdp/README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -207,7 +207,7 @@ Run point-goal evaluation:

```bash
bash eval/scripts/run_evaluation.sh \
--config_file eval/config/eval_pointgoal/quadruped_clutter_easy.yaml
--config_file eval/config/eval_pointgoal/quadruped_internscene_commercial.yaml
```

By default, evaluation runs all start-goal samples in the selected scene. Use `--num_episodes` to evaluate a subset and `--max_steps` to cap the total simulation steps for smoke tests or debugging.
Expand Down
10 changes: 5 additions & 5 deletions baselines/x-navdp/eval/config/eval_pointgoal/humanoid_clutter_easy.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
run_id: default
run_id: pointgoal_humanoid_eval_clutter_easy
run_root_dir: outputs/evaluation
logger: null

Expand All @@ -7,10 +7,10 @@ environment:
task_type: pointgoal
scene_dir: data/scenes/cluttered_easy
scene_index: 0
num_envs: 4
num_envs: 1
rgb_history_window: 1024
depth_history_window: 1
rgb_num_samples: 8
rgb_num_samples: 7
depth_num_samples: 1
rgb_image_size: 224
depth_image_size: 224
Expand Down Expand Up @@ -43,9 +43,9 @@ obs_mapping:
robot_quat: robot_rot

mpc:
N: 30 # Prediction horizon
N: 75 # Prediction horizon
desired_v: 0.5 # Desired velocity
v_max: 0.5 # Maximum velocity
w_max: 0.5 # Maximum angular velocity
ref_gap: 3 # Reference gap
T: 0.1 # Time step
T: 0.04 # Time step
10 changes: 5 additions & 5 deletions baselines/x-navdp/eval/config/eval_pointgoal/humanoid_clutter_hard.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
run_id: default
run_id: pointgoal_humanoid_eval_clutter_hard
run_root_dir: outputs/evaluation
logger: null

Expand All @@ -7,10 +7,10 @@ environment:
task_type: pointgoal
scene_dir: data/scenes/cluttered_hard
scene_index: 0
num_envs: 4
num_envs: 1
rgb_history_window: 1024
depth_history_window: 1
rgb_num_samples: 8
rgb_num_samples: 7
depth_num_samples: 1
rgb_image_size: 224
depth_image_size: 224
Expand Down Expand Up @@ -43,9 +43,9 @@ obs_mapping:
robot_quat: robot_rot

mpc:
N: 30 # Prediction horizon
N: 75 # Prediction horizon
desired_v: 0.5 # Desired velocity
v_max: 0.5 # Maximum velocity
w_max: 0.5 # Maximum angular velocity
ref_gap: 3 # Reference gap
T: 0.1 # Time step
T: 0.04 # Time step
12 changes: 6 additions & 6 deletions baselines/x-navdp/eval/config/eval_pointgoal/quadruped_clutter_easy.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
run_id: default
run_id: pointgoal_quadruped_eval_clutter_easy
run_root_dir: outputs/evaluation
logger: null

Expand All @@ -7,14 +7,14 @@ environment:
task_type: pointgoal
scene_dir: data/scenes/cluttered_easy
scene_index: 0
num_envs: 4
num_envs: 1
rgb_history_window: 1024
depth_history_window: 1
rgb_num_samples: 8
rgb_num_samples: 7
depth_num_samples: 1
rgb_image_size: 224
depth_image_size: 224
height_offset: 0.2
height_offset: 0.4
stuck_window_size: 50
stuck_threshold: 0.1
arrival_threshold: 1.0
Expand Down Expand Up @@ -43,9 +43,9 @@ obs_mapping:
robot_quat: robot_rot

mpc:
N: 30 # Prediction horizon
N: 75 # Prediction horizon
desired_v: 0.5 # Desired velocity
v_max: 0.5 # Maximum velocity
w_max: 0.5 # Maximum angular velocity
ref_gap: 3 # Reference gap
T: 0.1 # Time step
T: 0.04 # Time step
12 changes: 6 additions & 6 deletions baselines/x-navdp/eval/config/eval_pointgoal/quadruped_clutter_hard.yaml
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,4 +1,4 @@
run_id: default
run_id: pointgoal_quadruped_eval_clutter_hard
run_root_dir: outputs/evaluation
logger: null

Expand All @@ -7,14 +7,14 @@ environment:
task_type: pointgoal
scene_dir: data/scenes/cluttered_hard
scene_index: 0
num_envs: 4
num_envs: 1
rgb_history_window: 1024
depth_history_window: 1
rgb_num_samples: 8
rgb_num_samples: 7
depth_num_samples: 1
rgb_image_size: 224
depth_image_size: 224
height_offset: 0.2
height_offset: 0.4
stuck_window_size: 50
stuck_threshold: 0.1
arrival_threshold: 1.0
Expand Down Expand Up @@ -43,9 +43,9 @@ obs_mapping:
robot_quat: robot_rot

mpc:
N: 30 # Prediction horizon
N: 75 # Prediction horizon
desired_v: 0.5 # Desired velocity
v_max: 0.5 # Maximum velocity
w_max: 0.5 # Maximum angular velocity
ref_gap: 3 # Reference gap
T: 0.1 # Time step
T: 0.04 # Time step
2 changes: 1 addition & 1 deletion baselines/x-navdp/eval/scripts/run_evaluation.sh
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -45,7 +45,7 @@ done

if [ -z "$CONFIG_FILE" ]; then
echo "Error: --config_file is required"
echo "Usage: bash scripts/run_evaluation.sh --config_file eval/config/eval_pointgoal/quadruped_clutter_easy.yaml"
echo "Usage: bash scripts/run_evaluation.sh --config_file eval/config/eval_pointgoal/quadruped_internscene_commercial.yaml"
exit 1
fi

Expand Down
6 changes: 3 additions & 3 deletions baselines/x-navdp/src/environment/tasks/event_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -320,9 +320,9 @@ def pointnav_reset_eval(env: ManagerBasedEnv,
if global_occ_path is not None:
global _occ_cache_path, _occ_cache_data
if _occ_cache_path != global_occ_path:
_gs = 0.1 * scale * scale_factor
_gs = 0.1 * scale_factor
occ_pcd = o3d.io.read_point_cloud(global_occ_path)
occ_point = np.array(occ_pcd.points) * scale * scale_factor
occ_point = np.array(occ_pcd.points) * scale_factor
_occ_min, _occ_max = occ_point.min(axis=0), occ_point.max(axis=0)
grid_dimension = np.ceil((_occ_max[0:2] - _occ_min[0:2]) / _gs).astype(int)
decision_map = np.zeros(grid_dimension, dtype=np.int32)
Expand Down Expand Up @@ -363,7 +363,7 @@ def pointnav_reset_eval(env: ManagerBasedEnv,
for i in range(env_ids.shape[0]):
idx = int(count % sample_points.shape[0])
start_goal_pair = sample_points[idx].copy()
start_goal_pair[:4] = start_goal_pair[:4] * scale
start_goal_pair[:4] = start_goal_pair[:4] * scale_factor
start_points = np.array([start_goal_pair[0], start_goal_pair[1], 0])
goal_points = np.array([start_goal_pair[2], start_goal_pair[3], 0])

Expand Down