RAFA-Net: Region Attention Network For Food Items And Agricultural Stress Recognition

TL;DR

The proposed region attention network for food items and agricultural stress recognition method, dubbed RAFA-Net, has been experimented on three public food datasets, and has achieved state-of-the-art performances with distinct margins, implying RAFA-Net's generalization capability.

Abstract

Deep Convolutional Neural Networks (CNNs) have facilitated remarkable success in recognizing various food items and agricultural stress. A decent performance boost has been witnessed in solving the agro-food challenges by mining and analyzing of region-based partial feature descriptors. Also, computationally expensive ensemble learning schemes using multiple CNNs have been studied in earlier works. This work proposes a region attention scheme for modelling long-range dependencies by building a correlation among different regions within an input image. The attention method enhances feature representation by learning the usefulness of context information from complementary regions. Spatial pyramidal pooling and average pooling pair aggregate partial descriptors into a holistic representation. Both pooling methods establish spatial and channel-wise relationships without incurring extra parameters. A context gating scheme is applied to refine the descriptiveness of weighted attentional features, which is relevant for classification. The proposed Region Attention network for Food items and Agricultural stress recognition method, dubbed RAFA-Net, has been experimented on three public food datasets, and has achieved state-of-the-art performances with distinct margins. The highest top-1 accuracies of RAFA-Net are 91.69%, 91.56%, and 96.97% on the UECFood-100, UECFood-256, and MAFood-121 datasets, respectively. In addition, better accuracies have been achieved on two benchmark agricultural stress datasets. The best top-1 accuracies on the Insect Pest (IP-102) and PlantDoc-27 plant disease datasets are 92.36%, and 85.54%, respectively; implying RAFA-Net's generalization capability.

Figures (7)

• Figure 1: The proposed RAFA-Net consists of a region-pooling module upon which attention is applied for enhancing distinctness of various regions. The spatial and channel-wise relationships among attentioned regions are captured by feed-forward paths, and the features are further refined for classification.
• Figure 2: Visual variations in samples of different datasets, shown column-wise. Top row shows visual similarities among (a) food dish, (b) plant disease, and (c) insect pest classes. Bottom row reflects their structural and texture variances in natural environment, implies underlying recognition challenges.
• Figure 3: The proposed RAFA-Net consists of two sub-stages. a) Feature map pooling from regions of the same size. b) Region attention technique and further refinement by the feed-forward networks in two paths, which are combined together based on attention weights before the classification layer.
• Figure 4: Class-wise image distributions of the UCEFood100, MAFood121, PlantDoc27, and IP102 datasets. The curve of UCEFood256 is akin to UCEFood100.
• Figure 5: Sample dishes from the UECF-100& 256, and MAF-121 datasets