# Lightweight Distillation of SAM 3 and DINOv3 for Edge-Deployable Individual-Level Livestock Monitoring and Longitudinal Visual Analytics

arXiv: https://arxiv.org/abs/2604.27128 · HTML: https://mlsanity.com/paper/2604.27128/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 24 of its 50 references that are in the map.
- First posted on arXiv 29 Apr 2026.

## Builds on (most cited first, 24 of 24)

- [Adam: A Method for Stochastic Optimization](https://mlsanity.com/paper/1412.6980/) (2014)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Segment Anything](https://mlsanity.com/paper/2304.02643/) (2023)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift](https://mlsanity.com/paper/1502.03167/) (2015)
- [SGDR: Stochastic Gradient Descent with Warm Restarts](https://mlsanity.com/paper/1608.03983/) (2016)
- [Feature Pyramid Networks for Object Detection](https://mlsanity.com/paper/1612.03144/) (2016)
- [Gaussian Error Linear Units (GELUs)](https://mlsanity.com/paper/1606.08415/) (2016)
- [SAM 2: Segment Anything in Images and Videos](https://mlsanity.com/paper/2408.00714/) (2024)
- [Group Normalization](https://mlsanity.com/paper/1803.08494/) (2018)
- [Accurate, Large Minibatch SGD: Training ImageNet in 1 Hour](https://mlsanity.com/paper/1706.02677/) (2017)
- [DINOv3](https://mlsanity.com/paper/2508.10104/) (2025)
- [Mixed Precision Training](https://mlsanity.com/paper/1710.03740/) (2017)
- [SAM 3: Segment Anything with Concepts](https://mlsanity.com/paper/2511.16719/) (2025)
- [Performance Measures and a Data Set for Multi-Target, Multi-Camera Tracking](https://mlsanity.com/paper/1609.01775/) (2016)
- [MOT16: A Benchmark for Multi-Object Tracking](https://mlsanity.com/paper/1603.00831/) (2016)
- [Faster Segment Anything: Towards Lightweight SAM for Mobile Applications](https://mlsanity.com/paper/2306.14289/) (2023)
- [Scaling Open-Vocabulary Object Detection](https://mlsanity.com/paper/2306.09683/) (2023)
- [ReZero is All You Need: Fast Convergence at Large Depth](https://mlsanity.com/paper/2003.04887/) (2020)
- [TinyViT: Fast Pretraining Distillation for Small Vision Transformers](https://mlsanity.com/paper/2207.10666/) (2022)
- [Fixup Initialization: Residual Learning Without Normalization](https://mlsanity.com/paper/1901.09321/) (2019)
- [EfficientSAM: Leveraged Masked Image Pretraining for Efficient Segment Anything](https://mlsanity.com/paper/2312.00863/) (2023)
- [A Computer Vision Pipeline for Individual-Level Behavior Analysis: Benchmarking on the Edinburgh Pig Dataset](https://mlsanity.com/paper/2509.12047/) (2025)
- [EfficientSAM3: Progressive Hierarchical Distillation for Video Concept Segmentation from SAM1, 2, and 3](https://mlsanity.com/paper/2511.15833/) (2025)

## Abstract (arXiv)

Foundation-model pipelines for individual-level livestock monitoring -- combining open-vocabulary detection, promptable video segmentation, and self-supervised visual embeddings -- have raised the accuracy ceiling of precision livestock farming (PLF), but their GPU memory budgets exceed the envelope of commodity edge accelerators. To close this gap, the 446M-parameter Perception Encoder (PE-ViT-L+) backbone of SAM 3 is distilled into a 40.66M-parameter multi-scale student through three mechanisms: a Feature Pyramid Network student encoder built on TinyViT-21M-512, a four-term direction-then-scale distillation loss, and backbone-substitution inference with sliding-window session pruning that bounds streaming GPU memory growth. The DINOv3 family includes a pre-distilled ViT-S/16 variant (21.6M parameters) released alongside a 6716M-parameter ViT-7B teacher; the ViT-S (21M) variant is adopted as the per-individual embedder. On the Edinburgh Pig dataset, the compressed pipeline reaches 92.29% MOTA and 96.15% IDF1 against the SAM 3 teacher (1.68- and 0.84-percentage-point losses), achieves a 7.77-fold reduction in system-level parameters and a 3.01-fold reduction in peak VRAM (19.52GB -> 6.49GB), and reaches 97.34% top-1 accuracy with 91.67% macro-F1 on nine-class pig behaviour classification. The pipeline fits inside an NVIDIA Jetson Orin NX 16GB envelope with 4.9GB of headroom, supporting a proposed -- but not yet empirically validated -- on-device embedding-pool re-identification mechanism whose per-individual footprint of approximately 94MB per animal per year produces a longitudinal visual record amenable to retrospective association with disease, lameness, reproductive, and growth outcome labels.