Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Fusion Complexity Inversion: Why Simpler Cross View Modules Outperform SSMs and Cross View Attention Transformers for Pasture Biomass Regression

Mridankan Mandal

TL;DR

Adaptation of vision foundation models to agricultural regression is systematically evaluated and actionable guidelines for sparse agricultural benchmarks are established: backbone quality should be prioritized over fusion complexity, local modules preferred over global alternatives, and features unavailable at inference excluded.

Abstract

Accurate estimation of pasture biomass from agricultural imagery is critical for sustainable livestock management, yet existing methods are limited by the small, imbalanced, and sparsely annotated datasets typical of real world monitoring. In this study, adaptation of vision foundation models to agricultural regression is systematically evaluated on the CSIRO Pasture Biomass benchmark, a 357 image dual view dataset with laboratory validated, component wise ground truth for five biomass targets, through 17 configurations spanning four backbones (EfficientNet-B3 to DINOv3-ViT-L), five cross view fusion mechanisms, and a 4x2 metadata factorial. A counterintuitive principle, termed "fusion complexity inversion", is uncovered: on scarce agricultural data, a two layer gated depthwise convolution (R^2 = 0.903) outperforms cross view attention transformers (0.833), bidirectional SSMs (0.819), and full Mamba (0.793, below the no fusion baseline). Backbone pretraining scale is found to monotonically dominate all architectural choices, with the DINOv2 -> DINOv3 upgrade alone yielding +5.0 R^2 points. Training only metadata (species, state, and NDVI) is shown to create a universal ceiling at R^2 ~ 0.829, collapsing an 8.4 point fusion spread to 0.1 points. Actionable guidelines for sparse agricultural benchmarks are established: backbone quality should be prioritized over fusion complexity, local modules preferred over global alternatives, and features unavailable at inference excluded.

Fusion Complexity Inversion: Why Simpler Cross View Modules Outperform SSMs and Cross View Attention Transformers for Pasture Biomass Regression

TL;DR

Adaptation of vision foundation models to agricultural regression is systematically evaluated and actionable guidelines for sparse agricultural benchmarks are established: backbone quality should be prioritized over fusion complexity, local modules preferred over global alternatives, and features unavailable at inference excluded.

Abstract

Accurate estimation of pasture biomass from agricultural imagery is critical for sustainable livestock management, yet existing methods are limited by the small, imbalanced, and sparsely annotated datasets typical of real world monitoring. In this study, adaptation of vision foundation models to agricultural regression is systematically evaluated on the CSIRO Pasture Biomass benchmark, a 357 image dual view dataset with laboratory validated, component wise ground truth for five biomass targets, through 17 configurations spanning four backbones (EfficientNet-B3 to DINOv3-ViT-L), five cross view fusion mechanisms, and a 4x2 metadata factorial. A counterintuitive principle, termed "fusion complexity inversion", is uncovered: on scarce agricultural data, a two layer gated depthwise convolution (R^2 = 0.903) outperforms cross view attention transformers (0.833), bidirectional SSMs (0.819), and full Mamba (0.793, below the no fusion baseline). Backbone pretraining scale is found to monotonically dominate all architectural choices, with the DINOv2 -> DINOv3 upgrade alone yielding +5.0 R^2 points. Training only metadata (species, state, and NDVI) is shown to create a universal ceiling at R^2 ~ 0.829, collapsing an 8.4 point fusion spread to 0.1 points. Actionable guidelines for sparse agricultural benchmarks are established: backbone quality should be prioritized over fusion complexity, local modules preferred over global alternatives, and features unavailable at inference excluded.
Paper Structure (21 sections, 3 equations, 13 figures, 8 tables)

This paper contains 21 sections, 3 equations, 13 figures, 8 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Method
  4. Problem Formulation
  5. Dual View Input Pipeline
  6. Architecture
  7. Baseline Fusion Mechanisms
  8. Metadata Injection (Ablation Variant)
  9. Experiments
  10. Dataset
  11. Cross-Validation Protocol
  12. Training Configuration
  13. Main Results
  14. Fusion Complexity Analysis
  15. The Metadata Paradox
  16. ...and 6 more sections

Figures (13)

  • Figure 1: Architecture overview of the proposed dual-view biomass regression pipeline. Each input image is split into left/right halves, encoded by a weight-tied DINOv3-ViT-L backbone, and fused through two stacked GatedDepthwiseConv blocks before compositional prediction heads.
  • Figure 2: Histograms of the five biomass target variables in the training set ($n = 357$).
  • Figure 3: Pearson correlation heatmap among biomass targets and metadata variables.
  • Figure 4: NDVI and compressed height vs. biomass scatter plots, colored by pasture species.
  • Figure 5: Dry Total biomass distributions by Australian state.
  • ...and 8 more figures