Divide and Ensemble: Progressively Learning for the Unknown

TL;DR

The paper introduces DEEM, a progressive learning method that partitions data by collection dates, uses ensemble pseudo-labeling, and unifies models to improve classification accuracy in nutrient deficiency detection.

Contribution

It proposes a novel progressive learning approach with data partitioning and ensemble pseudo-labeling, achieving state-of-the-art results in nutrient deficiency classification.

Findings

Achieved 93.6% Top-1 accuracy on test data.

Outperformed other methods in the nutrient deficiency challenge.

Demonstrated effectiveness of date-based data partitioning and ensemble pseudo-labeling.

Abstract

In the wheat nutrient deficiencies classification challenge, we present the DividE and EnseMble (DEEM) method for progressive test data predictions. We find that (1) test images are provided in the challenge; (2) samples are equipped with their collection dates; (3) the samples of different dates show notable discrepancies. Based on the findings, we partition the dataset into discrete groups by the dates and train models on each divided group. We then adopt the pseudo-labeling approach to label the test data and incorporate those with high confidence into the training set. In pseudo-labeling, we leverage models ensemble with different architectures to enhance the reliability of predictions. The pseudo-labeling and ensembled model training are iteratively conducted until all test samples are labeled. Finally, the separated models for each group are unified to obtain the model for the whole dataset. Our method achieves an average of 93.6\% Top-1 test accuracy~(94.0\% on WW2020 and 93.2\% on WR2021) and wins the 1$st$ place in the Deep Nutrient Deficiency Challenge~\footnote{https://cvppa2023.github.io/challenges/}.