Visualizations

[caoyunkang]

I observed issues with numerical precision across different GPUs, which can result in poor visualizations in the original code. This is primarily due to the anomaly scores being exceedingly small, around 1e-6 . To address this, it is recommended to use the following revised code for tools/visualization.py to improve the visualizations.

def plot_sample_cv2(names, imgs, scores_: dict, gts, save_folder=None):
    os.makedirs(save_folder, exist_ok=True)

    # get subplot number
    total_number = len(imgs)

    scores = copy.deepcopy(scores_)
    # normarlisze anomalies
    for k, v in scores.items():
        max_value = np.max(v)
        min_value = np.min(v)

        # scores[k] = (scores[k] - min_value) / max_value * 255
        # scores[k] = scores[k].astype(np.uint8)

        # scale_factor = 255 / (max_value - min_value)
        # scores[k] = ((scores[k] - min_value) * scale_factor).round().clip(0, 255).astype(np.uint8)

        for i in range(len(scores[k])):
            cur_max_value = np.percentile(scores[k][i], 98)
            cur_min_value = np.percentile(scores[k][i], 2)

            scale_factor = 255 / (cur_max_value - cur_min_value)
            scores[k][i] = ((scores[k][i] - cur_min_value) * scale_factor).round().clip(0, 255).astype(np.uint8)

    # draw gts
    mask_imgs = []
    for idx in range(total_number):
        gts_ = gts[idx]
        mask_imgs_ = imgs[idx].copy()
        mask_imgs_[gts_ > 0.5] = (255, 255, 255)
        mask_imgs.append(mask_imgs_)

    # save imgs
    for idx in range(total_number):

        cv2.imwrite(os.path.join(save_folder, f'{names[idx]}_img.png'), imgs[idx])

        mask_img = cv2.addWeighted(mask_imgs[idx], 0.5, imgs[idx], 0.5, 0)
        cv2.imwrite(os.path.join(save_folder, f'{names[idx]}_mask.png'), mask_img)
        # cv2.imwrite(os.path.join(save_folder, f'{names[idx]}_mask.jpg'), mask_img)

        for key in scores:
            heat_map = cv2.applyColorMap(scores[key][idx], cv2.COLORMAP_JET)
            visz_map = cv2.addWeighted(heat_map, 0.5, imgs[idx], 0.5, 0)
            cv2.imwrite(os.path.join(save_folder, f'{names[idx]}_{key}.png'),
                        visz_map)