def parameters():
params = TrackerParams()
params.debug = 0
params.visualization = False
params.use_gpu = True
params.image_sample_size = 22 * 16
params.search_area_scale = 6
params.border_mode = 'inside_major'
params.patch_max_scale_change = 1.5
# Learning parameters
params.sample_memory_size = 50
params.learning_rate = 0.01
params.init_samples_minimum_weight = 0.25
params.train_skipping = 20
# Net optimization params
params.update_classifier = True
params.net_opt_iter = 10
params.net_opt_update_iter = 2
params.net_opt_hn_iter = 1
# Detection parameters
params.window_output = False
# Init augmentation parameters
params.use_augmentation = True
params.augmentation = {
'fliplr': True,
'rotate': [10, -10, 45, -45],
'blur': [(3, 1), (1, 3), (2, 2)],
'relativeshift': [(0.6, 0.6), (-0.6, 0.6), (0.6, -0.6), (-0.6, -0.6)],
'dropout': (2, 0.2)
}
params.augmentation_expansion_factor = 2
params.random_shift_factor = 1 / 3
# Advanced localization parameters
params.advanced_localization = True
params.target_not_found_threshold = 0.25
params.distractor_threshold = 0.8
params.hard_negative_threshold = 0.45
params.target_neighborhood_scale = 2.2
params.dispalcement_scale = 0.8
params.hard_negative_learning_rate = 0.02
params.update_scale_when_uncertain = True
# IoUnet parameters
params.box_refinement_space = 'relative'
params.iounet_augmentation = False # Use the augmented samples to compute the modulation vector
params.iounet_k = 3 # Top-k average to estimate final box
params.num_init_random_boxes = 9 # Num extra random boxes in addition to the classifier prediction
params.box_jitter_pos = 0.1 # How much to jitter the translation for random boxes
params.box_jitter_sz = 0.5 # How much to jitter the scale for random boxes
params.maximal_aspect_ratio = 6 # Limit on the aspect ratio
params.box_refinement_iter = 10 # Number of iterations for refining the boxes
params.box_refinement_step_length = 2.5e-3 # 1 # Gradient step length in the bounding box refinement
params.box_refinement_step_decay = 1 # Multiplicative step length decay (1 means no decay)
params.net = NetWithBackbone(net_path='super_dimp.pth.tar',
use_gpu=params.use_gpu)
params.vot_anno_conversion_type = 'preserve_area'
params.perform_hn_without_windowing = False
params.save_sample_interval = 2 # save memory interval
# [------ new parameters -------]
# parameters for re-detection
params.re_detection = True # [default:True] Re-detection
params.flag_confidence = 6 # [default:6] different methods about confidence score
params.cnt_global = 10000 # [default:10000->not use] global search for each count
params.global_search_memory_limit = 200 # global search memory limit
params.cnt_random = 5 # [default:5] random search for each count
params.additional_candidate_random = 3 # [default:3] the number of candidates to search (when additional_candidate_adaptive is False)
params.additional_candidate_adaptive = True # [default:True] adaptive number of candidates to search
params.additional_candidate_adaptive_ratio = 0.1 # [default:0.1] ratio for adaptive number
params.additional_candidate_adaptive_min = 1 # [default:1] minimum number for searching
params.additional_candidate_adaptive_max = 10 # [default:10] maximum number for searching
params.redetection_score_penalty = True # [default:True] score penalty for re-detection
params.redetection_score_penalty_alpha = 0.75 # [default:0.75] score penalty parameter, about distance (big value -> more penalty)
params.redetection_score_penalty_beta = 0.25 # [default:0.25] score penalty parameter, about time (small value -> slow detect)
params.redetection_basic_penalty = 0.75 # [default:0.75] score penalty parameter, total score reduction, (0.75 -> 25% reduce score)
params.redetection_now = True # [default:True] re-detection immediately after tracking failure
params.no_update_early_redetection = 1 # no update period after re-detection success
params.no_save_early_redetection = 0 # no save period after re-detection success
params.redetection_global_search_flag = 1 # search position flag (0: 1/4 overlap, 1: half overlap, 2: none overlap)
# parameters for more discriminative learning
params.init_net_more_learn = True # more discriminative learning (init)
params.init_more_learn_expand_searching_size = True # expanding searching size for more init samples
params.init_more_learn_flag = 1 # search position flag (0: 1/4 overlap, 1: half overlap, 2: none overlap)
params.init_more_learn_memory_limit = 40 # 100 -> 4Gb (smaller than sample_memory_size - 2), default:40
params.init_more_learn_no_transform = True # whether transformation
params.init_more_sample_limit = 10 # searching number limit
params.init_iou_more_learn = True # same as init_net_more_learn, iounet_augmentation = True -> can change
params.track_net_more_learn = True # more discriminative learning (track)
params.track_net_more_learn_search_flag = 2 # search position flag (0: 1/4 overlap, 1: half overlap, 2: none overlap)
params.track_net_more_learn_cnt = 3 # more learning period
params.track_net_more_learn_score = 0.70 # score condition for more learning
params.track_net_more_learn_not_save = True # if True: not save and learn each time (not depend on track_net_more_learn_cnt)
params.track_net_more_learn_save_weight = 0.5 # reduce memory weight (due to not real data)
params.train_more_sample_limit = 10 # searching number limit
params.additional_train_candidate = 2 # number of additional samples
params.memory_weight_ratio = 2 # weight for initial feature (higher -> more important)
params.init_blending = 0.0 # image blending with target and background (0: not blending, 0.1: 10% background, 90% target)
# parameters for random erasing
params.erasing_mode = True # Random erasing (RE) mode when tracking
params.use_original_pos = False # Random erasing flag
params.erasing_cnt = 5 # Random erasing period
params.lower_scale = 0.02 # RE parameters
params.upper_scale = 0.05 # RE parameters
params.lower_ratio = 0.7 # RE parameters
params.upper_ratio = 1.3 # RE parameters
params.num_erasing = 10 # the number of random erasing images
return params
def parameters():
params = TrackerParams()
# These are usually set from outside
params.debug = 0 # Debug level
params.visualization = False # Do visualization
# Use GPU or not (IoUNet requires this to be True)
params.use_gpu = True
# Feature specific parameters
deep_params = TrackerParams()
# Patch sampling parameters
params.max_image_sample_size = (18 * 16)**2 # Maximum image sample size
params.min_image_sample_size = (18 * 16)**2 # Minimum image sample size
params.search_area_scale = 5 # Scale relative to target size
params.feature_size_odd = False # Good to use False for even-sized kernels and vice versa
# Optimization parameters
params.CG_iter = 5 # The number of Conjugate Gradient iterations in each update after the first frame
params.init_CG_iter = 60 # The total number of Conjugate Gradient iterations used in the first frame
params.init_GN_iter = 6 # The number of Gauss-Newton iterations used in the first frame (only if the projection matrix is updated)
params.post_init_CG_iter = 0 # CG iterations to run after GN
params.fletcher_reeves = False # Use the Fletcher-Reeves (true) or Polak-Ribiere (false) formula in the Conjugate Gradient
params.standard_alpha = True # Use the standard formula for computing the step length in Conjugate Gradient
params.CG_forgetting_rate = None # Forgetting rate of the last conjugate direction
# Learning parameters for each feature type
deep_params.learning_rate = 0.01 # Learning rate
deep_params.init_samples_minimum_weight = 0.25 # Minimum weight of initial samples in memory
deep_params.output_sigma_factor = 1 / 4 # Standard deviation of Gaussian label relative to target size
# Training parameters
params.sample_memory_size = 250 # Memory size
params.train_skipping = 10 # How often to run training (every n-th frame)
# Online model parameters
deep_params.kernel_size = (4, 4) # Kernel size of filter
deep_params.compressed_dim = 64 # Dimension output of projection matrix
deep_params.filter_reg = 1e-1 # Filter regularization factor
deep_params.projection_reg = 1e-4 # Projection regularization factor
# Windowing
params.feature_window = False # Perform windowing of features
params.window_output = False # Perform windowing of output scores
# Detection parameters
params.scale_factors = torch.ones(
1
) # What scales to use for localization (only one scale if IoUNet is used)
params.score_upsample_factor = 1 # How much Fourier upsampling to use
# Init data augmentation parameters
params.augmentation = {
'fliplr': True,
'rotate': [5, -5, 10, -10, 20, -20, 30, -30, 45, -45, -60, 60],
'blur': [(2, 0.2), (0.2, 2), (3, 1), (1, 3), (2, 2)],
'relativeshift': [(0.6, 0.6), (-0.6, 0.6), (0.6, -0.6), (-0.6, -0.6)],
'dropout': (7, 0.2)
}
params.augmentation_expansion_factor = 2 # How much to expand sample when doing augmentation
params.random_shift_factor = 1 / 3 # How much random shift to do on each augmented sample
deep_params.use_augmentation = True # Whether to use augmentation for this feature
# Factorized convolution parameters
# params.use_projection_matrix = True # Use projection matrix, i.e. use the factorized convolution formulation
params.update_projection_matrix = True # Whether the projection matrix should be optimized or not
params.proj_init_method = 'randn' # Method for initializing the projection matrix
params.filter_init_method = 'randn' # Method for initializing the spatial filter
params.projection_activation = 'none' # Activation function after projection ('none', 'relu', 'elu' or 'mlu')
params.response_activation = (
'mlu', 0.05
) # Activation function on the output scores ('none', 'relu', 'elu' or 'mlu')
# Advanced localization parameters
params.advanced_localization = True # Use this or not
params.target_not_found_threshold = 0.25 # Absolute score threshold to detect target missing
params.distractor_threshold = 0.8 # Relative threshold to find distractors
params.hard_negative_threshold = 0.5 # Relative threshold to find hard negative samples
params.target_neighborhood_scale = 2.2 # Target neighborhood to remove
params.dispalcement_scale = 0.8 # Dispacement to consider for distractors
params.hard_negative_learning_rate = 0.02 # Learning rate if hard negative detected
params.hard_negative_CG_iter = 5 # Number of optimization iterations to use if hard negative detected
params.update_scale_when_uncertain = True # Update scale or not if distractor is close
# IoUNet parameters
params.use_iou_net = True # Use IoU net or not
params.box_refinement_space = 'relative'
params.iounet_augmentation = False # Use the augmented samples to compute the modulation vector
params.iounet_k = 3 # Top-k average to estimate final box
params.num_init_random_boxes = 9 # Num extra random boxes in addition to the classifier prediction
params.box_jitter_pos = 0.1 # How much to jitter the translation for random boxes
params.box_jitter_sz = 0.5 # How much to jitter the scale for random boxes
params.maximal_aspect_ratio = 6 # Limit on the aspect ratio
params.box_refinement_iter = 10 # Number of iterations for refining the boxes
params.box_refinement_step_length = (
1e-2, 5e-2
) # 1 # Gradient step length in the bounding box refinement 5e-3 2e-2
params.box_refinement_step_decay = 1 # Multiplicative step length decay (1 means no decay)
# Setup the feature extractor (which includes the IoUNet)
deep_fparams = FeatureParams(feature_params=[deep_params])
deep_feat = deep.ATOMResNet18(net_path='atom_gmm_sampl',
output_layers=['layer3'],
fparams=deep_fparams,
normalize_power=2)
params.features = MultiResolutionExtractor([deep_feat])
return params
def parameters():
params = TrackerParams()
params.debug = 0
params.visualization = False
params.use_gpu = True
params.image_sample_size = 14*16
params.search_area_scale = 4
params.border_mode = 'inside_major'
params.patch_max_scale_change = 1.5
# Learning parameters
params.sample_memory_size = 250
params.learning_rate = 0.0075
params.init_samples_minimum_weight = 0.0
params.train_skipping = 10
# Net optimization params
params.update_classifier = True
params.net_opt_iter = 25
params.net_opt_update_iter = 3
params.net_opt_hn_iter = 3
# Detection parameters
params.window_output = True
# Init augmentation parameters
params.use_augmentation = True
params.augmentation = {'fliplr': True,
'rotate': [-5, 10, -30, 60],
'blur': [(2, 0.2), (1, 3)],
'relativeshift': [(0.6, 0.6), (-0.6, -0.6)],
'dropout': (3, 0.2)}
params.augmentation_expansion_factor = 2
params.random_shift_factor = 1/3
# Advanced localization parameters
params.advanced_localization = True
params.target_not_found_threshold = 0.0
params.distractor_threshold = 100
params.hard_negative_threshold = 0.45
params.target_neighborhood_scale = 2.2
params.dispalcement_scale = 0.7
params.perform_hn_without_windowing = True
params.hard_negative_learning_rate = 0.02
params.update_scale_when_uncertain = True
# IoUnet parameters
params.box_refinement_space = 'relative'
params.iounet_augmentation = False # Use the augmented samples to compute the modulation vector
params.iounet_k = 3 # Top-k average to estimate final box
params.num_init_random_boxes = 9 # Num extra random boxes in addition to the classifier prediction
params.box_jitter_pos = 0.1 # How much to jitter the translation for random boxes
params.box_jitter_sz = 0.5 # How much to jitter the scale for random boxes
params.maximal_aspect_ratio = 6 # Limit on the aspect ratio
params.box_refinement_iter = 10 # Number of iterations for refining the boxes
params.box_refinement_step_length = 2.5e-3 # 1 # Gradient step length in the bounding box refinement
params.box_refinement_step_decay = 1 # Multiplicative step length decay (1 means no decay)
params.net = NetWithBackbone(net_path='trdimp_net.pth.tar', use_gpu=params.use_gpu)
params.vot_anno_conversion_type = 'preserve_area'
return params
