def re_ranking(probFea, galFea, k1, k2, lambda_value, local_distmat=None, only_local=False):
"""
probFea: all feature vectors of the query set (torch tensor)
galFea: all feature vectors of the gallery set (torch tensor)
k1,k2,lambda: parameters, the original paper uses (k1=20,k2=6,lambda=0.3)
Code adapted from https://github.com/zhunzhong07/person-re-ranking
Zhong Z, Zheng L, Cao D, et al. Re-ranking Person Re-identification with k-reciprocal Encoding CVPR 2017.
"""
# if feature vector is numpy, you should use 'torch.tensor' transform it to tensor
query_num = probFea.size(0)
all_num = query_num + galFea.size(0)
if only_local:
original_dist = local_distmat
else:
feat = torch.cat([probFea, galFea])
# print('using GPU to compute original distance')
distmat = (
torch.pow(feat, 2).sum(dim=1, keepdim=True).expand(all_num, all_num)
+ torch.pow(feat, 2).sum(dim=1, keepdim=True).expand(all_num, all_num).t()
)
distmat.addmm_(1, -2, feat, feat.t())
original_dist = distmat.cpu().numpy()
del feat
if local_distmat is not None:
original_dist = original_dist + local_distmat
gallery_num = original_dist.shape[0]
original_dist = np.transpose(original_dist / np.max(original_dist, axis=0))
V = np.zeros_like(original_dist).astype(np.float16)
initial_rank = np.argsort(original_dist).astype(np.int32)
# print('starting re_ranking')
for i in range(all_num):
# k-reciprocal neighbors
forward_k_neigh_index = initial_rank[i, : k1 + 1]
backward_k_neigh_index = initial_rank[forward_k_neigh_index, : k1 + 1]
fi = np.where(backward_k_neigh_index == i)[0]
k_reciprocal_index = forward_k_neigh_index[fi]
k_reciprocal_expansion_index = k_reciprocal_index
for j in range(len(k_reciprocal_index)):
candidate = k_reciprocal_index[j]
candidate_forward_k_neigh_index = initial_rank[candidate, : int(np.around(k1 / 2)) + 1]
candidate_backward_k_neigh_index = initial_rank[
candidate_forward_k_neigh_index, : int(np.around(k1 / 2)) + 1
]
fi_candidate = np.where(candidate_backward_k_neigh_index == candidate)[0]
candidate_k_reciprocal_index = candidate_forward_k_neigh_index[fi_candidate]
if len(np.intersect1d(candidate_k_reciprocal_index, k_reciprocal_index)) > 2 / 3 * len(
candidate_k_reciprocal_index
):
k_reciprocal_expansion_index = np.append(k_reciprocal_expansion_index, candidate_k_reciprocal_index)
k_reciprocal_expansion_index = np.unique(k_reciprocal_expansion_index)
weight = np.exp(-original_dist[i, k_reciprocal_expansion_index])
V[i, k_reciprocal_expansion_index] = weight / np.sum(weight)
original_dist = original_dist[:query_num,]
if k2 != 1:
V_qe = np.zeros_like(V, dtype=np.float16)
for i in range(all_num):
V_qe[i, :] = np.mean(V[initial_rank[i, :k2], :], axis=0)
V = V_qe
del V_qe
del initial_rank
invIndex = []
for i in range(gallery_num):
invIndex.append(np.where(V[:, i] != 0)[0])
jaccard_dist = np.zeros_like(original_dist, dtype=np.float16)
for i in range(query_num):
temp_min = np.zeros(shape=[1, gallery_num], dtype=np.float16)
indNonZero = np.where(V[i, :] != 0)[0]
indImages = [invIndex[ind] for ind in indNonZero]
for j in range(len(indNonZero)):
temp_min[0, indImages[j]] = temp_min[0, indImages[j]] + np.minimum(
V[i, indNonZero[j]], V[indImages[j], indNonZero[j]]
)
jaccard_dist[i] = 1 - temp_min / (2 - temp_min)
final_dist = jaccard_dist * (1 - lambda_value) + original_dist * lambda_value
del original_dist
del V
del jaccard_dist
final_dist = final_dist[:query_num, query_num:]
return final_dist