Autoregressive Adaptive Hypergraph Transformer for Skeleton-based Activity Recognition

Extracting multiscale contextual information and higher-order correlations among skeleton sequences using Graph Convolutional Networks (GCNs) alone is inadequate for effective action classification. Hypergraph convolution addresses the above issues but cannot harness the long-range dependencies. Transformer proves to be effective in capturing these dependencies and making complex contextual features accessible. We propose an Autoregressive Adaptive HyperGraph Transformer (AutoregAd-HGformer) model for in-phase (autoregressive and discrete) and out-phase (adaptive) hypergraph generation. The vector quantized in-phase hypergraph equipped with powerful autoregressive learned priors produces a more robust and informative representation suitable for hyperedge formation. The out-phase hypergraph generator provides a model-agnostic hyperedge learning technique to align the attributes with input skeleton embedding. The hybrid (supervised and unsupervised) learning in AutoregAd-HGformer explores the action-dependent feature along spatial, temporal, and channel dimensions. The extensive experimental results and ablation study indicate the superiority of our model over state-of-the-art hypergraph architectures on NTU RGB+D, NTU RGB+D 120, and NW-UCLA datasets.

HWRCNet: Handwritten Word Recognition in JPEG Compressed Domain using CNN-BiLSTM Network

The handwritten word recognition from images using deep learning is an active research area with promising performance. It practical scenario, it might be required to process the handwritten images in the compressed domain due to due to security reasons. However, the utilization of deep learning is still very limited for the processing of compressed images. Motivated by the need of processing document images in the compressed domain using recent developments in deep learning, we propose a HWRCNet model for handwritten word recognition in JPEG compressed domain. The proposed model combines the Convolutional Neural Network (CNN) and Bi-Directional Long Short Term Memory (BiLSTM) based Recurrent Neural Network (RNN). Basically, we train the model using compressed domain images and observe a very appealing performance with 89.05% word recognition accuracy and 13.37% character error rate.