Gait recognition is a biometric modality that identifies individuals from their characteristic walking patterns. Unlike conventional biometric traits, gait can be acquired at a distance and without active subject cooperation, making it suitable for surveillance and public safety applications. Nevertheless, silhouette-based temporal models remain sensitive to long sequences, observation noise, and appearance-related covariates. Recurrent architectures often struggle to preserve information from earlier frames and are inherently sequential to optimize, whereas transformer-based models typically require greater computational resources and larger training sets and may be sensitive to irregular sequence lengths and noisy inputs. These limitations reduce robustness under clothing variation, carrying conditions, and view changes, while also hindering the joint modeling of local gait cycles and longer-term motion trends. To address these challenges, we introduce a Temporal Kolmogorov-Arnold Network (TKAN) for gait recognition. The proposed model replaces fixed edge weights with learnable one-dimensional functions and incorporates a two-level memory mechanism consisting of short-term RKAN sublayers and a gated long-term pathway. This design enables efficient modeling of both cycle-level dynamics and broader temporal context while maintaining a compact backbone. Experiments on the CASIA-B dataset indicate that the proposed CNN+TKAN framework achieves strong recognition performance under the reported evaluation setting.
Figures & tables
Figure 1: Proposed pipeline. A shared CNN encodes each silhouette frame into a compact feature; the frame features are stacked and batch-normalized to form a sequence. The TKAN head models short- and long-range gait dynamics over the sequence and outputs refined time-step features, which are temporally averaged (with dropout) to obtain a clip-level descriptor. A linear–softmax classifier maps this descriptor to the identity posterior and the top-1 label is reported. (In our setup: T=50 , feature width d=256 , classes C=74 .)
Figure 2: Architecture of CNN model
Figure 3: Training/validation loss and accuracy for CNN+TKAN on CASIA–B (subject split 1–74 train, 75–124 test).
Method
NM
BG
CL
Avg
CNN+TKAN (Ours)
99.52
99.56
98.82
99.30
VPNet (CVPR’24) [ 12 ]
98.3
96.3
90.0
94.9
CLTD (ECCV’24) [ 13 ]
98.6
96.4
89.3
94.8
HSTL (ICCV’23) [ 14 ]
98.1
95.9
88.9
94.3
GaitGCI (CVPR’23) [ 15 ]
98.4
96.6
88.5
94.5
DyGait (ICCV’23) [ 16 ]
98.4
96.2
87.8
94.1
Table 1: Comparison on CASIA-B (Rank-1 %) under the standard protocol (subjects 1–74 train, 75–124 test; identical-view excluded). † indicates enlarged resolution (e.g., 128 × 88).
Method
NM
BG
CL
CNN + LSTM
98.06 / 99.63
96.76 / 99.56
92.56 / 98.90
CNN + Transformer
98.86 / 99.82
97.50 / 99.85
95.21 / 98.45
CNN + TKAN (ours)
99.52 / 100.00
99.56 / 100.00
98.82 / 100.00
Table 2: Internal comparison (same CNN, training, and protocol). Rank-1 / Rank-5 (%) on CASIA-B by condition.
Figure 4: ROC and AUC for CNN+TKAN on CASIA–B, showing both micro- and macro-averaged one-vs-rest ROC curves, reported separately for NM, BG, and CL.
Gait recognition, as a promising biometric technology, identifies individuals through their unique walking patterns and offers distinctive advantages including non-invasiveness, long-range applicability, and resistance to deliberate disguise. Despite these merits, capturing the intrinsic motion patterns concealed within consecutive video frames remains challenging due to the complexity of video data and the interference of external covariates such as viewpoint changes, clothing variations, and carrying conditions. Existing approaches predominantly rely on either static appearance features extracted from individual silhouette frames or employ complex sequential models (\eg, LSTM, 3D convolutions) that demand substantial computational resources and sophisticated training strategies. To address these limitations, we propose a Local Spatiotemporal Convolutional Network (LSTCN), a structurally simple yet highly effective dual-branch architecture that endows standard two-dimensional convolutional networks with the capacity to extract temporal information. Specifically, we introduce a Global Bidirectional Spatial Pooling (GBSP) mechanism that reduces the dimensionality of gait tensors by decomposing spatial features into horizontal and vertical strip-based local representations, enabling the temporal dimension to participate in standard 2D convolution operations. Building upon this, we design a Local Spatiotemporal Convolutional (LSTC) layer that jointly processes temporal and spatial dimensions, allowing the network to adaptively learn strip-based gait motion patterns. We further extend this formulation with asymmetric convolution kernels that independently attend to the temporal, spatial, and joint spatiotemporal domains, thereby enriching the extracted feature representations.
Xiaoyun Wang, Cunrong Li, Wu Wang
School of Mechanical and Electrical Engineering, Osh State University
Gait recognition has emerged as a compelling biometric modality for surveillance and security applications, offering inherent advantages such as non-intrusiveness, resistance to disguise, and long-range identification capability. However, prevailing approaches struggle to comprehensively capture and exploit the rich biometric cues embedded in human locomotion, particularly under covariate interference including viewpoint variation, clothing change, and carrying conditions. In this paper, we present a high-precision gait recognition framework that deeply extracts and synergistically fuses gait dynamics with body shape characteristics through a multi-branch architecture grounded in deep residual learning. Specifically, we first employ the High-Resolution Network (HRNet) to perform robust skeletal keypoint estimation, preserving fine-grained spatial information even under low-resolution inputs. We then construct three complementary feature branches -- body proportion, gait velocity, and skeletal motion -- from the extracted pose sequences. A 50-layer Residual Network (ResNet-50) backbone is leveraged within a deep feature extraction module to capture hierarchically rich and discriminative representations. To effectively integrate heterogeneous feature streams, we design a Multi-Branch Feature Fusion (MFF) module inspired by channel-wise attention mechanisms, which dynamically allocates contribution weights across branches through learned activation parameters. Extensive experiments on the cross-view multi-condition CASIA-B benchmark demonstrate that our method achieves a Rank-1 accuracy of 94.52% under normal walking, with the best recognition performance among skeleton-based methods for the coat-wearing condition.
Yabo Luo, Xiaoyun Wang, Cunrong Li
School of Mechanical and Electrical Engineering, Osh State University
3D skeleton-based gait emotion recognition faces high annotation costs, data scarcity, and poor generalization on heterogeneous data. This paper proposes SV-GCN, a single-stream multi-feature fusion framework with temporal invariance. We introduce intra-frame relative motion features to eliminate frame-rate sensitivity and embed heterogeneous cues at shallow layers, enabling early fusion without multi-stream complexity. For variable-length sequences, we design a global mask-guided valid-frame spatio-temporal graph convolution module, introducing frame-rate insensitivity for the first time in this domain. On the E-Gait dataset, our method achieves performance comparable to state-of-the-art while demonstrating strong generalization across varying sequence lengths and frame rates, offering a viable pathway for pre-training on large-scale skeleton-based action recognition datasets.
Shirong Lyu, Silu Quan, Yixuan Ding +1
College of Computer and Information Science, Southwest University, Chongqing, China · College of Physical Science and Technology, Southwest University, Chongqing, China · Wisesoft Inc., Chengdu, China