Pose estimation has been a hot topic in the field of machine vision in recent years. In the pose estimation task, a lightweight stacked hourglass network (SHN) algorithm is proposed. Moreover, aiming at the problem of large parameters in depthwise convolutional neural networks, a lightweight residual module is proposed, that is, based on the lightweight efficient channel attention improved conditional channel-weighted method (ICCW-Bottle), which replaces bottleneck module, thereby reducing the weight of the network and obtaining the feature information of different scales. Given the problem that a large amount of feature information is easily lost after the network pooling operation, a lightweight dual-branch fusion module is proposed that fully integrates high-level semantic information and low-level detailed features under the condition of a small number of parameters. Finally, the training model of synthetic animal dataset and real animal dataset was jointly applied. Compared with the consistency-constrained semi-supervised learning (CC-SSL) method, the proposed method increased in accuracy of pose estimation by 5.5%. It also reduced the number of network parameters and the calculation amount. The results of the ablation experiment verify the advancement and effectiveness of the overall network.
Pose estimation has been a hot topic in the field of machine vision in recent years. Animals exist widely in nature, and the analysis of their shape and movement is important in many fields and industries. In the pose estimation task, to improve the detection accuracy, the existing models often need to consume a lot of computing and memory resources. Therefore, it is a key problem for the pose estimation methods to carry out a lightweight model and reduce the computational overhead on the premise of ensuring model accuracy. In this paper, we focus on the structure of the convolutional neural network in animal pose estimation, construct a lightweight and efficient stacked hourglass network model oriented to optimize the balance of model computation and accuracy, and implement the application algorithm design based on it. Aiming at the problem of large parameters in depthwise convolutional neural networks, a lightweight residual module is proposed, that is, based on the lightweight efficient channel attention improved conditional channel-weighted method (ICCW-Bottle), thereby reducing the weight of the network and obtaining the feature information of different scales. Given the problem that a large amount of feature information is easily lost after the network pooling operation, a lightweight dual-branch fusion module is proposed that fully integrates high-level semantic information and low-level detailed features under the condition of a small number of parameters. Finally, the same as the CC-SSL method: the model is trained jointly using synthetic and real animal datasets, but the CC-SSL method does not take into account the computational power of the model, which consumes a lot of time and memory to run. Through experiments, it is known that compared with the CC-SSL method, the PCK @ 0.05 of this method is increased by 5.5 % on the TigDog dataset. The model in this paper reduces the number of parameters and calculations of the network while ensuring less information loss and model accuracy. The ablation experiment verifies the advancement and effectiveness of the overall network.INDEX TERMS Animal pose estimation, stacked Hourglass Networks, lightweight, residual module, feature fusion.
C28H41NO2, orthorhombic, P212121 (no. 19), a = 10.4480(5) Å, b = 12.4188(6) Å, c = 18.3905(10) Å, V = 2386.2(2) Å3, Z = 4, R
gt
(F) = 0.0608, wR
ref
(F
2) = 0.1510, T = 273(2) K.
C28H41NO, triclinic,
P
1
$P1$
(no. 1), a = 6.3144(8) Å, b = 9.7315(13) Å, c = 10.5867(14) Å, α = 111.561(4)°, β =
96.486
(
4
)
∘
$96.486{(4)}^{\circ }$
, γ = 97.128(4)°, V = 591.41(14) Å3, Z = 1,
R
g
t
${R}_{\mathit{g}\mathit{t}}$
(F) = 0.0579,
w
R
r
e
f
$w{R}_{\mathit{r}\mathit{e}\mathit{f}}$
(F
2) = 0.1687, T = 273(2) K.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.