Computational Intelligence and Neuroscience / 2019 / Article / Tab 5

Research Article

Triplet Deep Hashing with Joint Supervised Loss Based on Deep Neural Networks

Table 5

MAP results for different numbers of bits (12, 24, 32, and 48 bits) on the two benchmark image datasets (CIFAR-10 and NUS-WIDE).

MethodsCIFAR-10NUS-WIDE
12 bits24 bits32 bits48 bits12 bits24 bits32 bits48 bits

Deep methodsOurs0.7700.7740.7880.7830.7910.8280.8320.836
DTSH0.7100.7500.7650.7740.7730.8080.8120.824
DPSH0.7130.7270.7440.7570.7520.7900.7940.812
DQN0.5540.5580.5640.5800.7680.7760.7830.792
DHN0.5550.5940.6030.6210.7080.7350.7480.758
NINH0.5520.5660.5580.5810.6740.6970.7130.715
CNNH0.4390.5110.5090.5220.6110.6180.6250.608

Nondeep methodsFastH0.3050.3490.3690.3840.6210.6500.6650.687
SDH0.2850.3290.3410.3560.5680.6000.6080.637
KSH0.3030.3370.3460.3560.5560.5720.5810.588
LFH0.1760.2310.2110.2530.5710.5680.5680.585
SPLH0.1710.1730.1780.1840.5680.5890.5970.601
ITQ0.1620.1690.1720.1750.4520.4680.4720.477
SH0.1270.1280.1260.1290.4540.4060.4050.400

We are committed to sharing findings related to COVID-19 as quickly and safely as possible. Any author submitting a COVID-19 paper should notify us at help@hindawi.com to ensure their research is fast-tracked and made available on a preprint server as soon as possible. We will be providing unlimited waivers of publication charges for accepted articles related to COVID-19. Sign up here as a reviewer to help fast-track new submissions.