def _read(self, feat_path, label_path, proposal_folders):
with Timer('read meta and feature'):
if label_path is not None:
self.lb2idxs, self.idx2lb = read_meta(label_path)
self.labels = intdict2ndarray(self.idx2lb)
self.inst_num = len(self.idx2lb)
self.ignore_label = False
else:
self.lb2idxs, self.idx2lb = None, None
self.labels = None
self.inst_num = -1
self.ignore_label = True
if not self.featureless:
features = read_probs(feat_path, self.inst_num,
self.feature_dim)
self.features = l2norm(features)
if self.inst_num == -1:
self.inst_num = features.shape[0]
else:
assert self.inst_num > 0
self.feature_dim = 1
self.features = np.ones(self.inst_num).reshape(-1, 1)
with Timer('read proposal list'):
self.lst = []
self.tot_lst = []
if callable(proposal_folders):
proposal_folders = proposal_folders()
for proposal_folder in proposal_folders:
print('read proposals from folder: ', proposal_folder)
fn_nodes = sorted(
glob.glob(osp.join(proposal_folder, self.fn_node_pattern)))
fn_edges = sorted(
glob.glob(osp.join(proposal_folder, self.fn_edge_pattern)))
assert len(fn_nodes) == len(
fn_edges), "node files({}) vs edge files({})".format(
len(fn_nodes), len(fn_edges))
assert len(fn_nodes) > 0, 'files under {} is 0'.format(
proposal_folder)
for fn_node, fn_edge in zip(fn_nodes, fn_edges):
# sanity check
assert fn_node[:fn_node.rfind(
'_')] == fn_edge[:fn_edge.rfind('_'
)], "{} vs {}".format(
fn_node, fn_edge)
if self._check_iop(fn_node):
self.lst.append([fn_node, fn_edge])
self.tot_lst.append([fn_node, fn_edge])
self.size = len(self.lst)
self.tot_size = len(self.tot_lst)
assert self.size <= self.tot_size
if self.size < self.tot_size:
print('select {} / {} = {:.2f} proposals '
'with iop between ({:.2f}, {:.2f})'.format(
self.size, self.tot_size,
1. * self.size / self.tot_size, self.th_iop_min,
self.th_iop_max))
def test_lgcn(model, cfg, logger):
for k, v in cfg.model['kwargs'].items():
setattr(cfg.test_data, k, v)
dataset = build_dataset(cfg.test_data)
ofn_pred = os.path.join(cfg.work_dir, 'pred_edges_scores.npz')
if os.path.isfile(ofn_pred) and not cfg.force:
data = np.load(ofn_pred)
edges = data['edges']
scores = data['scores']
inst_num = data['inst_num']
if inst_num != len(dataset):
logger.warn(
'instance number in {} is different from dataset: {} vs {}'.
format(ofn_pred, inst_num, len(dataset)))
else:
edges, scores, inst_num = test(model, dataset, cfg, logger)
# produce predicted labels
clusters = graph_clustering_dynamic_th(edges,
scores,
max_sz=cfg.max_sz,
step=cfg.step,
pool=cfg.pool)
pred_idx2lb = clusters2labels(clusters)
pred_labels = intdict2ndarray(pred_idx2lb)
if cfg.save_output:
print('save predicted edges and scores to {}'.format(ofn_pred))
np.savez_compressed(ofn_pred,
edges=edges,
scores=scores,
inst_num=inst_num)
ofn_meta = os.path.join(cfg.work_dir, 'pred_labels.txt')
write_meta(ofn_meta, pred_idx2lb, inst_num=inst_num)
# evaluation
if not dataset.ignore_label:
print('==> evaluation')
gt_labels = dataset.labels
for metric in cfg.metrics:
evaluate(gt_labels, pred_labels, metric)
single_cluster_idxs = get_cluster_idxs(clusters, size=1)
print('==> evaluation (removing {} single clusters)'.format(
len(single_cluster_idxs)))
remain_idxs = np.setdiff1d(np.arange(len(dataset)),
np.array(single_cluster_idxs))
remain_idxs = np.array(remain_idxs)
for metric in cfg.metrics:
evaluate(gt_labels[remain_idxs], pred_labels[remain_idxs], metric)
def _read(self, feat_path, label_path, proposal_folders):
fn_node_pattern = '*_node.npz'
fn_edge_pattern = '*_edge.npz'
with Timer('read meta and feature'):
if label_path is not None:
self.lb2idxs, self.idx2lb = read_meta(label_path)
self.labels = intdict2ndarray(self.idx2lb)
self.inst_num = len(self.idx2lb)
self.ignore_label = False
else:
self.lb2idxs, self.idx2lb = None, None
self.labels = None
self.inst_num = -1
self.ignore_label = True
if not self.featureless:
features = read_probs(feat_path, self.inst_num,
self.feature_dim)
self.features = l2norm(features)
if self.inst_num == -1:
self.inst_num = features.shape[0]
else:
assert self.inst_num > 0
self.feature_dim = 1
self.features = np.ones(self.inst_num).reshape(-1, 1)
with Timer('read proposal list'):
self.lst = []
if callable(proposal_folders):
proposal_folders = proposal_folders()
for proposal_folder in proposal_folders:
print('read proposals from folder: ', proposal_folder)
fn_nodes = sorted(
glob.glob(os.path.join(proposal_folder, fn_node_pattern)))
fn_edges = sorted(
glob.glob(os.path.join(proposal_folder, fn_edge_pattern)))
assert len(fn_nodes) == len(
fn_edges), "node files({}) vs edge files({})".format(
len(fn_nodes), len(fn_edges))
assert len(fn_nodes) > 0, 'files under {} is 0'.format(
proposal_folder)
for fn_node, fn_edge in zip(fn_nodes, fn_edges):
assert fn_node[:fn_node.rfind(
'_')] == fn_edge[:fn_edge.rfind('_'
)], "{} vs {}".format(
fn_node, fn_edge)
self.lst.append([fn_node, fn_edge])
self.size = len(self.lst)
def __init__(self, cfg):
feat_path = cfg['feat_path']
label_path = cfg.get('label_path', None)
knn_graph_path = cfg['knn_graph_path']
self.k_at_hop = cfg['k_at_hop']
self.depth = len(self.k_at_hop)
self.active_connection = cfg['active_connection']
self.feature_dim = cfg['feature_dim']
self.is_norm_feat = cfg.get('is_norm_feat', True)
self.is_sort_knns = cfg.get('is_sort_knns', True)
self.is_test = cfg.get('is_test', False)
with Timer('read meta and feature'):
if label_path is not None:
_, idx2lb = read_meta(label_path)
self.inst_num = len(idx2lb)
self.labels = intdict2ndarray(idx2lb)
self.ignore_label = False
else:
self.labels = None
self.inst_num = -1
self.ignore_label = True
self.features = read_probs(feat_path, self.inst_num,
self.feature_dim)
if self.is_norm_feat:
self.features = l2norm(self.features)
if self.inst_num == -1:
self.inst_num = self.features.shape[0]
self.size = self.inst_num
with Timer('read knn graph'):
knns = np.load(knn_graph_path)['data']
_, self.knn_graph = knns2ordered_nbrs(knns, sort=self.is_sort_knns)
assert np.mean(self.k_at_hop) >= self.active_connection
print('feature shape: {}, norm_feat: {}, sort_knns: {} '
'k_at_hop: {}, active_connection: {}'.format(
self.features.shape, self.is_norm_feat, self.is_sort_knns,
self.k_at_hop, self.active_connection))
def __init__(self, cfg):
feat_path = cfg['feat_path']
label_path = cfg.get('label_path', None)
knn_graph_path = cfg.get('knn_graph_path', None)
self.k = cfg['k']
self.feature_dim = cfg['feature_dim']
self.is_norm_feat = cfg.get('is_norm_feat', True)
self.th_sim = cfg.get('th_sim', 0.)
self.max_conn = cfg.get('max_conn', 1)
self.ignore_ratio = cfg.get('ignore_ratio', 0.8)
self.ignore_small_confs = cfg.get('ignore_small_confs', True)
self.use_candidate_set = cfg.get('use_candidate_set', True)
self.nproc = cfg.get('nproc', 1)
self.max_qsize = cfg.get('max_qsize', int(1e5))
with Timer('read meta and feature'):
if label_path is not None:
self.lb2idxs, self.idx2lb = read_meta(label_path)
self.inst_num = len(self.idx2lb)
self.gt_labels = intdict2ndarray(self.idx2lb)
self.ignore_label = False
else:
self.inst_num = -1
self.ignore_label = True
self.features = read_probs(feat_path, self.inst_num,
self.feature_dim)
if self.is_norm_feat:
self.features = l2norm(self.features)
if self.inst_num == -1:
self.inst_num = self.features.shape[0]
self.size = self.inst_num
assert self.size == self.features.shape[0]
print('feature shape: {}, k: {}, norm_feat: {}'.format(
self.features.shape, self.k, self.is_norm_feat))
with Timer('read knn graph'):
if knn_graph_path is not None:
knns = np.load(knn_graph_path)['data']
else:
prefix = osp.dirname(feat_path)
name = rm_suffix(osp.basename(feat_path))
# find root folder of `features`
prefix = osp.dirname(prefix)
knn_prefix = osp.join(prefix, 'knns', name)
knns = build_knns(knn_prefix, self.features, cfg.knn_method,
cfg.knn)
assert self.inst_num == len(knns), "{} vs {}".format(
self.inst_num, len(knns))
adj = fast_knns2spmat(knns, self.k, self.th_sim, use_sim=True)
# build symmetric adjacency matrix
adj = build_symmetric_adj(adj, self_loop=True)
self.adj = row_normalize(adj)
# convert knns to (dists, nbrs)
self.dists, self.nbrs = knns2ordered_nbrs(knns, sort=True)
if cfg.pred_confs != '':
print('read estimated confidence from {}'.format(
cfg.pred_confs))
self.confs = np.load(cfg.pred_confs)['pred_confs']
else:
print('use unsupervised density as confidence')
assert self.radius
from vegcn.confidence import density
self.confs = density(self.dists, radius=self.radius)
assert 0 <= self.ignore_ratio <= 1
if self.ignore_ratio == 1:
self.ignore_set = set(np.arange(len(self.confs)))
else:
num = int(len(self.confs) * self.ignore_ratio)
confs = self.confs
if not self.ignore_small_confs:
confs = -confs
self.ignore_set = set(np.argpartition(confs, num)[:num])
print(
'ignore_ratio: {}, ignore_small_confs: {}, use_candidate_set: {}'.
format(self.ignore_ratio, self.ignore_small_confs,
self.use_candidate_set))
print('#ignore_set: {} / {} = {:.3f}'.format(
len(self.ignore_set), self.inst_num,
1. * len(self.ignore_set) / self.inst_num))
with Timer('Prepare sub-graphs'):
# construct subgraphs with larger confidence
self.peaks = {i: [] for i in range(self.inst_num)}
self.dist2peak = {i: [] for i in range(self.inst_num)}
if self.nproc > 1:
# multi-process
import multiprocessing as mp
pool = mp.Pool(self.nproc)
results = []
num = int(self.inst_num / self.max_qsize) + 1
for i in tqdm(range(num)):
beg = int(i * self.max_qsize)
end = min(beg + self.max_qsize, self.inst_num)
lst = [j for j in range(beg, end)]
results.extend(
list(
tqdm(pool.map(self.get_subgraph, lst),
total=len(lst))))
pool.close()
pool.join()
else:
results = [
self.get_subgraph(i) for i in tqdm(range(self.inst_num))
]
self.adj_lst = []
self.feat_lst = []
self.lb_lst = []
self.subset_gt_labels = []
self.subset_idxs = []
self.subset_nbrs = []
self.subset_dists = []
for result in results:
if result is None:
continue
elif len(result) == 3:
i, nbr, dist = result
self.peaks[i].extend(nbr)
self.dist2peak[i].extend(dist)
continue
i, nbr, dist, feat, adj, lb = result
self.subset_idxs.append(i)
self.subset_nbrs.append(nbr)
self.subset_dists.append(dist)
self.feat_lst.append(feat)
self.adj_lst.append(adj)
if not self.ignore_label:
self.subset_gt_labels.append(self.idx2lb[i])
self.lb_lst.append(lb)
self.subset_gt_labels = np.array(self.subset_gt_labels)
self.size = len(self.feat_lst)
assert self.size == len(self.adj_lst)
if not self.ignore_label:
assert self.size == len(self.lb_lst)
def __init__(self, cfg):
feat_path = cfg['feat_path']
label_path = cfg.get('label_path', None)
knn_graph_path = cfg.get('knn_graph_path', None)
self.k = cfg['k']
self.feature_dim = cfg['feature_dim']
self.is_norm_feat = cfg.get('is_norm_feat', True)
self.save_decomposed_adj = cfg.get('save_decomposed_adj', False)
self.th_sim = cfg.get('th_sim', 0.)
self.max_conn = cfg.get('max_conn', 1)
self.conf_metric = cfg.get('conf_metric')
with Timer('read meta and feature'):
if label_path is not None:
self.lb2idxs, self.idx2lb = read_meta(label_path)
self.inst_num = len(self.idx2lb)
self.gt_labels = intdict2ndarray(self.idx2lb)
self.ignore_label = False
else:
self.inst_num = -1
self.ignore_label = True
self.features = read_probs(feat_path, self.inst_num,
self.feature_dim)
if self.is_norm_feat:
self.features = l2norm(self.features)
if self.inst_num == -1:
self.inst_num = self.features.shape[0]
self.size = 1 # take the entire graph as input
with Timer('read knn graph'):
if os.path.isfile(knn_graph_path):
knns = np.load(knn_graph_path)['data']
else:
if knn_graph_path is not None:
print('knn_graph_path does not exist: {}'.format(
knn_graph_path))
prefix = osp.dirname(feat_path)
name = rm_suffix(osp.basename(feat_path))
# find root folder of `features`
prefix = osp.dirname(prefix)
knn_prefix = osp.join(prefix, 'knns', name)
knns = build_knns(knn_prefix, self.features, cfg.knn_method,
cfg.knn)
adj = fast_knns2spmat(knns, self.k, self.th_sim, use_sim=True)
# build symmetric adjacency matrix
adj = build_symmetric_adj(adj, self_loop=True)
adj = row_normalize(adj)
if self.save_decomposed_adj:
adj = sparse_mx_to_indices_values(adj)
self.adj_indices, self.adj_values, self.adj_shape = adj
else:
self.adj = adj
# convert knns to (dists, nbrs)
self.dists, self.nbrs = knns2ordered_nbrs(knns)
print('feature shape: {}, k: {}, norm_feat: {}'.format(
self.features.shape, self.k, self.is_norm_feat))
if not self.ignore_label:
with Timer('Prepare ground-truth label'):
self.labels = confidence(feats=self.features,
dists=self.dists,
nbrs=self.nbrs,
metric=self.conf_metric,
idx2lb=self.idx2lb,
lb2idxs=self.lb2idxs)
if cfg.eval_interim:
_, self.peaks = confidence_to_peaks(
self.dists, self.nbrs, self.labels, self.max_conn)
def __init__(self, cfg):
feat_path = cfg['feat_path']
label_path = cfg.get('label_path', None)
knn_graph_path = cfg.get('knn_graph_path', None)
self.k = cfg['k']
self.feature_dim = cfg['feature_dim']
self.is_norm_feat = cfg.get('is_norm_feat', True)
self.save_decomposed_adj = cfg.get('save_decomposed_adj', False)
self.th_sim = cfg.get('th_sim', 0.)
self.max_conn = cfg.get('max_conn', 1)
self.conf_metric = cfg.get('conf_metric')
self.num_process = cfg.get('num_process',16)
with Timer('read meta and feature'):
if label_path is not None:
self.lb2idxs, self.idx2lb = read_meta(label_path)
self.inst_num = len(self.idx2lb)
self.gt_labels = intdict2ndarray(self.idx2lb)
self.ignore_label = False
else:
self.inst_num = -1
self.ignore_label = True
self.features = read_probs(feat_path, self.inst_num,
self.feature_dim)
if self.is_norm_feat:
self.features = l2norm(self.features)
if self.inst_num == -1:
self.inst_num = self.features.shape[0]
self.size = 1 # take the entire graph as input
with Timer('read knn graph'):
if os.path.isfile(knn_graph_path):
knns = np.load(knn_graph_path)['data'] # num_imgs*2*k
else:
if knn_graph_path is not None:
print('knn_graph_path does not exist: {}'.format(
knn_graph_path))
knn_prefix = os.path.join(cfg.prefix, 'knns', cfg.name)
# 通过faiss实现k近邻搜索，此处作者faiss_gpu版本实现可能有问题，但faiss大规模在cpu上跑还是慢
# 当然faiss有针内存和计算速度方面的优化，PQ,IVF等，可参考faiss
knns = build_knns(knn_prefix, self.features, cfg.knn_method,
cfg.knn,self.num_process)
# 依据k近邻搜索结果构建邻接矩阵
adj = fast_knns2spmat(knns, self.k, self.th_sim, use_sim=True)
# build symmetric adjacency matrix
adj = build_symmetric_adj(adj, self_loop=True)
adj = row_normalize(adj)
if self.save_decomposed_adj:
adj = sparse_mx_to_indices_values(adj)
self.adj_indices, self.adj_values, self.adj_shape = adj
else:
self.adj = adj
# convert knns to (dists, nbrs)
self.dists, self.nbrs = knns2ordered_nbrs(knns) # num_imgs*k
print('feature shape: {}, k: {}, norm_feat: {}'.format(
self.features.shape, self.k, self.is_norm_feat))
if not self.ignore_label:
with Timer('Prepare ground-truth label'):
self.labels = confidence(feats=self.features,
dists=self.dists,
nbrs=self.nbrs,
metric=self.conf_metric,
idx2lb=self.idx2lb,
lb2idxs=self.lb2idxs)
if cfg.eval_interim:
_, self.peaks = confidence_to_peaks(
self.dists, self.nbrs, self.labels, self.max_conn)
def __init__(self, cfg):
feat_path = cfg['feat_path']
label_path = cfg.get('label_path', None)
knn_graph_path = cfg.get('knn_graph_path', None)
self.k = cfg['k']
self.feature_dim = cfg['feature_dim']
self.is_norm_feat = cfg.get('is_norm_feat', True)
self.save_decomposed_adj = cfg.get('save_decomposed_adj', False)
self.th_sim = cfg.get('th_sim', 0.)
self.conf_metric = cfg.get('conf_metric')
with Timer('read meta and feature'):
if label_path is not None:
self.lb2idxs, self.idx2lb = read_meta(label_path)
self.inst_num = len(self.idx2lb)
self.cls_num = len(self.lb2idxs)
self.gt_labels = intdict2ndarray(self.idx2lb)
self.ignore_label = False
else:
self.inst_num = -1
self.ignore_label = True
self.features = read_probs(feat_path, self.inst_num,
self.feature_dim)
if self.is_norm_feat:
self.features = l2norm(self.features)
if self.inst_num == -1:
self.inst_num = self.features.shape[0]
self.size = 1 # take the entire graph as input
with Timer('Compute center feature'):
self.center_fea = np.zeros((self.cls_num, self.features.shape[1]))
for i in range(self.cls_num):
self.center_fea[i] = np.mean(self.features[self.lb2idxs[i]], 0)
self.center_fea = l2norm(self.center_fea)
with Timer('read knn graph'):
if os.path.isfile(knn_graph_path):
print("load knns from the knn_path")
self.knns = np.load(knn_graph_path)['data']
else:
if knn_graph_path is not None:
print('knn_graph_path does not exist: {}'.format(
knn_graph_path))
knn_prefix = os.path.join(cfg.prefix, 'knns', cfg.name)
self.knns = build_knns(knn_prefix, self.features,
cfg.knn_method, cfg.knn)
adj = fast_knns2spmat(self.knns, self.k, self.th_sim, use_sim=True)
# build symmetric adjacency matrix
adj = build_symmetric_adj(adj, self_loop=True)
#print('adj before norm')
#print(adj)
adj = row_normalize(adj)
if self.save_decomposed_adj:
adj = sparse_mx_to_indices_values(adj)
self.adj_indices, self.adj_values, self.adj_shape = adj
else:
self.adj = adj
# convert knns to (dists, nbrs)
self.dists, self.nbrs = knns2ordered_nbrs(self.knns)
print('feature shape: {}, k: {}, norm_feat: {}'.format(
self.features.shape, self.k, self.is_norm_feat))