tink/nerfstudio
4
0
Fork 0
mirror of https://github.com/nerfstudio-project/nerfstudio.git synced 2026-11-11 02:24:38 +08:00
Code

base-metrics runs

Browse Source
This commit is contained in:
Frederik Warburg frwa@dtu.dk 2023-07-26 12:51:29 +00:00
parent 2c4fd33cc7
commit 4dab1ca601
1 changed files with 15 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

29
nerfstudio/scripts/metrics.py
View File

@ -78,9 +78,9 @@ class BaseMetrics:
ssim = float(ssim_module(x, x_gt)[0])
lpips = float(lpips_module(x, x_gt)[0])
metrics["psnr_list"].append(psnr)
metrics["ssim_list"].append(ssim)
metrics["lpips_list"].append(lpips)
metrics["psnr_list"] = metrics.get("psnr_list", []) + [psnr]
metrics["ssim_list"] = metrics.get("ssim_list", []) + [ssim]
metrics["lpips_list"] = metrics.get("lpips_list", []) + [lpips]
# save the images
rgb_gt = (rgb_gt * 255.0).cpu().numpy().astype(np.uint8)
@ -107,6 +107,7 @@ class BaseMetrics:
# write to a json file
metrics_filename = self.output_folder / f"{experiment_name}.json"
os.makedirs(self.output_folder, exist_ok=True)
with open(metrics_filename, "w") as f:
json.dump(metrics, f, indent=4)
@ -217,21 +218,21 @@ class NerfbusterMetrics(BaseMetrics):
normals_median = float(theta.median())
# the angle thresholds
metrics["normals_11.25"].append(float((theta < 11.25).float().mean()))
metrics["normals_22.5"].append(float((theta < 22.5).float().mean()))
metrics["normals_30"].append(float((theta < 30).float().mean()))
metrics["normals_11.25"] = metrics.get("normals_11.25", []) + [float((theta < 11.25).float().mean())]
metrics["normals_22.5"] = metrics.get("normals_22.5", []) + [float((theta < 22.5).float().mean())]
metrics["normals_30"] = metrics.get("normals_30", []) + [float((theta < 30).float().mean())]
# coverage
coverage = float(mask[..., 0].sum() / visibilty_mask.sum())
metrics["psnr_list"].append(psnr)
metrics["ssim_list"].append(ssim)
metrics["lpips_list"].append(lpips)
metrics["depth_list"].append(depth_mse)
metrics["disparity_list"].append(disparity)
metrics["normals_list"].append(normals_mse)
metrics["normals_median_list"].append(normals_median)
metrics["coverage_list"].append(coverage)
metrics["psnr_list"] = metrics.get("psnr_list", []) + [psnr]
metrics["ssim_list"] = metrics.get("ssim_list", []) + [ssim]
metrics["lpips_list"] = metrics.get("lpips_list", []) + [lpips]
metrics["depth_list"] = metrics.get("depth_list", []) + [depth_mse]
metrics["disparity_list"] = metrics.get("disparity_list", []) + [disparity]
metrics["normals_list"] = metrics.get("normals_list", []) + [normals_mse]
metrics["normals_median_list"] = metrics.get("normals_median_list", []) + [normals_median]
metrics["coverage_list"] = metrics.get("coverage_list", []) + [coverage]
# save the images
rgb_gt = (rgb_gt * 255.0).cpu().numpy().astype(np.uint8)