data points 0.002001935
feature space 0.001922464
input data 0.001793878
bit encoding 0.00154062
bit quantisation 0.001527893
variable bit 0.0015141220000000001
optimal bit 0.001492309
single bit 0.001486715
bit count 0.001480825
bit counts 0.001449571
noun similarity 0.001438249
overall bit 0.0014092010000000001
large number 0.001356851
bit budget 0.00133401
bit allocation 0.001328271
input feature 0.001315454
double bit 0.00129952
bit alloca 0.00129952
unsupervised method 0.001297974
random lsh 0.001282297
quantisation matrix 0.001267214
dimensional feature 0.0012530010000000001
similarity lists 0.00124325
inal feature 0.001234629
ture space 0.001230531
similarity computations 0.0012100980000000002
pled method 0.0011961089999999999
random hyperplanes 0.001162375
time retrieval 0.001137223
other projection 0.001129877
variable number 0.001108546
constraint matrix 0.001106322
random restarts 0.001099289
different regions 0.0010894470000000001
indicator matrix 0.001045717
approximate approach 0.001039496
different sym 0.001030013
other work 0.001021651
language processing 0.001007522
projected values 9.88656E-4
standard text 9.83784E-4
multiple bits 9.83718E-4
search technique 9.819059999999998E-4
toy problem 9.77915E-4
maximum number 9.69014E-4
feature 9.64496E-4
timization algorithm 9.60421E-4
space 9.57968E-4
encoding scheme 9.529639999999999E-4
uniform number 9.46061E-4
normal vector 9.456969999999999E-4
similarity 9.44508E-4
identical number 9.44259E-4
total number 9.40312E-4
quantisation scheme 9.40237E-4
objective function 9.3776E-4
method 9.28334E-4
binary encoding 9.24285E-4
quantisation work 9.229049999999999E-4
based search 9.22049E-4
exact approach 9.19695E-4
constant number 9.152850000000001E-4
ural language 9.15171E-4
negative pairs 9.01911E-4
neighbour search 8.980679999999999E-4
same regions 8.97073E-4
bour search 8.90827E-4
tion pairs 8.89592E-4
same shape 8.787879999999999E-4
memory requirements 8.747099999999999E-4
projection stage 8.674010000000001E-4
projection functions 8.564150000000001E-4
quantisation technique 8.55268E-4
cosine distance 8.52039E-4
positive pairs 8.51617E-4
approximate techniques 8.46034E-4
recent quantisation 8.437099999999999E-4
integer linear 8.3906E-4
binary representation 8.351529999999999E-4
negative pair 8.35068E-4
query document 8.313330000000001E-4
lsh hyperplane 8.21685E-4
normal vectors 8.215849999999999E-4
quantisation stage 8.20716E-4
work locality 8.18133E-4
previous work 8.119080000000001E-4
uniform quantisation 8.0819E-4
preserving quantisation 7.990569999999999E-4
neighbourhood structure 7.95439E-4
specified pairs 7.901729999999999E-4
retrieval accuracy 7.88211E-4
ing quality 7.88193E-4
binary integer 7.87521E-4
positive pair 7.84774E-4
quantisation thresholds 7.78988E-4
lsh hyperplanes 7.77212E-4
threshold positions 7.763959999999999E-4
driven threshold 7.75684E-4
matrix 7.74991E-4
compact representation 7.74284E-4
