codingwolfman

garbage characters about segmentation results

1

[{'image_id': 0, 'category_id': 1, 'segmentation': {'size': [2688, 2208], 'counts': 'clZU1Y1eb29H7H7J3M3M3N2M2N3M2O2M3L5L6J6I9D;F9G9H8G5N3M2N2N2N1O2O1N1O2N2N2M3N2M4L3L5M3L4M2N2N2O1N1O2O1N101N1O2N1O2N1N3N2M3M3N2N2N3M2N2O1N2N101N100O2N1O1O1N3M2N2O1O2N1O1O1O1O2O0O100O1O2O0O1O1O1N3M[JneM`1QZ2XN]fMc1aY2XNkfMd1SY2XNUgMe1jX2[N[gMa1dX2_N_gM`1_X2`NcgM_1\\X2aNfgM^1YX2bNigM^1VX2aNlgM^1SX2bNogM^1PX2bNQhM^1mW2bNUhM]1kW2bNWhM_1fW2aN\\hM_1cW2`N_hM`1_W2`NdhM_1ZW2aNihM^1UW2aNohM^1oV2aNViM\\1iV2cNZiM]1cV2cNaiM[1]V2fNeiMZ1YV2fNiiMZ1UV2fNmiMY1RV2gNPjMX1oU2hNRjMY1lU2hNUjMW1kU2hNWjMX1gU2hNZjMX1fU2hN[jMW1eU2hN\\jMX1dU2gN^jMX1bU2hN_jMW1aU2hNajMW1^U2iNdjMV1\\U2hNgjMW1YU2hNjjMV1VU2hNmjMW1SU2gNQkMW1oT2hNTkMV1lT2hNWkMW1hT2iNZkMV1fT2iN\\kMV1dT2iN^kMV1bT2jN_kMU1aT2jN`kMV1`T2iNbkMV1^T2jNbkMW1\\T2iNfkMV1YT2kNgkMV1XT2iNjkMV1UT2jNmkMV1QT2kNPlMU1nS2kNTlMU1iS2kNZlMU1dS2kN^lMU1_S2lNdlMT1XS2lNllMT1QS2kNRmMU1lR2kNWmMT1gR2lN[mMT1cR2lN_mMT1_R2lNcmMT1[R2lNfmMU1YR2jNimMV1UR2kNlmMT1TR2kNmmMU1RR2lNnmMT1RR2lNnmMT1RR2lNnmMU1PR2lNQnMS1oQ2lNRnMT1mQ2mNSnMS1mQ2mNSnMS1lQ2nNTnMR1kQ2oNVnMP1iQ2POXnMP1gQ2QOYnMo0fQ2ROZnMn0eQ2SO[nMm0dQ2TO]nMk0bQ2UO_nMk0`Q2VO`nMj0_Q2WOanMi0^Q2XOcnMg0]Q2XOdnMi0ZQ2WOhnMh0XQ2WOjnMi0TQ2WOnnMh0RQ2WOPoMi0oP2VOSoMi0mP2VOVoMi0hP2VO[oMj0dP2UO^oMk0aP2TOaoMl0]P2TOeoMl0ZP2SOhoMn0VP2QOkoMP1TP2POmoMQ1QP2nNQPNR1mo1oNSPNR1lo1mNVPNR1jo1nNVPNS1io1lNXPNU1fo1lN[PNS1eo1lN\\PNT1co1mN^PNR1bo1mN_PNT1_o1lNcPNS1\\o1mNePNS1Zo1mNhPNR1Wo1mNlPNR1So1nNnPNR1Po1nNSQNQ1ln1nNWQNQ1gn1PO\\QNo0bn1QO`QNn0_n1ROcQNm0\\n1SOgQNk0Xn1UOjQNj0Un1VOmQNi0Rn1XOnQNh0Rn1WOPRNh0Pn1XOPRNh0Pn1XOQRNh0nm1XORRNh0mm1YOSRNg0mm1YOSRNg0mm1YOTRNf0lm1YOURNg0km1YOURNg0km1YOURNh0jm1XOVRNh0jm1XOVRNh0jm1XOWRNg0im1YOWRNg0im1XOXRNh0hm1XOXRNi0gm1WOYRNi0gm1WOYRNi0gm1WOZRNh0em1YO[RNg0em1XO\\RNi0bm1WO`RNi0_m1WOaRNj0]m1VOeRNi0Zm1WOhRNj0Um1UOnRNk0Pm1UORSNk0ll1TOWSNl0fl1UO\\SNl0al1TObSNk0[l1VOhSNh0Wl1XOkSNh0Rl1YOQTNf0mk1[OTTNe0kk1ZOWTNe0hk1[OYTNf0ek1[O\\TNd0ck1\\O_TNd0`k1\\O`TNd0`k1\\O`TNd0`k1\\O`TNd0`k1\\O`TNd0`k1\\O`TNd0`k1\\O`TNd0_k1]OaTNc0_k1\\OcTNc0]k1]OcTNc0]k1]OcTNc0]k1]OcTNc0]k1]OcTNc0]k1]OcTNc0]k1]OcTNc0]k1

could you provide scripts for cityscape to pascal voc style?

2

i use the prepare_data.m in https://github.com/yuhuayc/da-faster-rcnn, and meet one error >> system(sprintf('find %s -name "*.png"', ... fullfile(source_data_dir,'leftImg8bit','train'))) 拒绝访问 - CITYSCAPES\LEFTIMG8BIT\TRAIN 找不到文件 - -NAME ans = 1

how to download foggy cityscape

7

i sent a letter to , but no anwser

problem about dataparallel

2

when i set batchsize=1, everything is ok, but when i turn to batchsize=2, i meet error Traceback (most recent call last): File "da_trainval_net.py", line 393, in tgt_im_data, tgt_im_info, tgt_gt_boxes, tgt_num_boxes,...

is the semantic loss right?

1

# Moving Centroid decay = self.decay s_centroid = (1-decay) * self.s_centroid + decay * current_s_centroid t_centroid = (1-decay) * self.t_centroid + decay * current_t_centroid semantic_loss = self.MSEloss(s_centroid, t_centroid) self.s_centroid =...