Open Yorionice1 opened 1 year ago
Yes, we do not revise the default test scale. I think the performance can be improved if the test scale is fixed.
Thanks for your reply. But I still can't get the accuracy reported in your paper (50.9), my result is around 45.2 on 4 GPUs with IMG_PER_BATCH_LABEL: 8, IMG_PER_BATCH_UNLABEL: 8, BASE_LR: 0.04. Rest of the settings are same as original config file uploaded by you.
MODEL:
META_ARCHITECTURE: "DAobjTwoStagePseudoLabGeneralizedRCNN" BACKBONE: NAME: "build_vgg_backbone" MASK_ON: False RESNETS: DEPTH: 101 PROPOSAL_GENERATOR: NAME: "PseudoLabRPN"
RPN: IN_FEATURES: ["vgg4"] ROI_HEADS: NAME: "StandardROIHeadsPseudoLab" LOSS: "CrossEntropy" # variant: "CrossEntropy" NUM_CLASSES: 8 IN_FEATURES: ["vgg4"] ROI_BOX_HEAD: NAME: "FastRCNNConvFCHead" NUM_FC: 2 POOLER_RESOLUTION: 7 SOLVER: LR_SCHEDULER_NAME: "WarmupTwoStageMultiStepLR" STEPS: (60000, 80000, 90000, 360000) FACTOR_LIST: (1, 1, 1, 1, 1) MAX_ITER: 100000 IMG_PER_BATCH_LABEL: 8 IMG_PER_BATCH_UNLABEL: 8 BASE_LR: 0.04 DATALOADER: SUP_PERCENT: 100.0 DATASETS: CROSS_DATASET: True TRAIN_LABEL: ("cityscapes_train",) TRAIN_UNLABEL: ("cityscapes_foggy_train",) TEST: ("cityscapes_foggy_val",) SEMISUPNET: Trainer: "ateacher" BBOX_THRESHOLD: 0.8 TEACHER_UPDATE_ITER: 1 BURN_UP_STEP: 20000 EMA_KEEP_RATE: 0.9996 UNSUP_LOSS_WEIGHT: 1.0 SUP_LOSS_WEIGHT: 0.5 DIS_TYPE: "vgg4" #["concate","p2","multi"] TEST: EVAL_PERIOD: 1000
[09/30 02:19:46] detectron2 INFO: Running with full config: CUDNN_BENCHMARK: false DATALOADER: ASPECT_RATIO_GROUPING: true FILTER_EMPTY_ANNOTATIONS: true NUM_WORKERS: 4 RANDOM_DATA_SEED: 0 RANDOM_DATA_SEED_PATH: dataseed/COCO_supervision.txt REPEAT_THRESHOLD: 0.0 SAMPLER_TRAIN: TrainingSampler SUP_PERCENT: 100.0 DATASETS: CROSS_DATASET: true PRECOMPUTED_PROPOSAL_TOPK_TEST: 1000 PRECOMPUTED_PROPOSAL_TOPK_TRAIN: 2000 PROPOSAL_FILES_TEST: [] PROPOSAL_FILES_TRAIN: [] TEST:
Thanks for your working. I found the test scale is 800, which is larger than it in other papers in this field.