SAM-TTT: Segment Anything Model via Reverse Parameter Configuration and Test-Time Training for Camouflaged Object Detection

2025-09-15

AI Summary

Existing SAM-based camouflaged object detection (COD) models neglect suppression of detrimental parameters that impair downstream semantic understanding. This work proposes Reverse SAM, the first framework to jointly integrate reverse parameter configuration and test-time training (TTT) into COD. Specifically, a Reverse SAM Parameter Configuration Module identifies and suppresses harmful parameters, while a lightweight T-Visioner module—featuring linear-complexity TTT layers, sequence modeling, and visual feature enhancement—dynamically refines SAM’s semantic representation capability at inference time without additional training. Evaluated on mainstream COD benchmarks (e.g., CAMO, CHAMELEON), Reverse SAM achieves state-of-the-art performance, with significant improvements in mFβ. This work establishes a novel paradigm for adapting foundation models like SAM to low-level vision tasks through parameter-aware, test-time adaptation.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Calibration & Uncertainty Quantification
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

This paper introduces a new Segment Anything Model (SAM) that leverages reverse parameter configuration and test-time training to enhance its performance on Camouflaged Object Detection (COD), named SAM-TTT. While most existing SAM-based COD models primarily focus on enhancing SAM by extracting favorable features and amplifying its advantageous parameters, a crucial gap is identified: insufficient attention to adverse parameters that impair SAM's semantic understanding in downstream tasks. To tackle this issue, the Reverse SAM Parameter Configuration Module is proposed to effectively mitigate the influence of adverse parameters in a train-free manner by configuring SAM's parameters. Building on this foundation, the T-Visioner Module is unveiled to strengthen advantageous parameters by integrating Test-Time Training layers, originally developed for language tasks, into vision tasks. Test-Time Training layers represent a new class of sequence modeling layers characterized by linear complexity and an expressive hidden state. By integrating two modules, SAM-TTT simultaneously suppresses adverse parameters while reinforcing advantageous ones, significantly improving SAM's semantic understanding in COD task. Our experimental results on various COD benchmarks demonstrate that the proposed approach achieves state-of-the-art performance, setting a new benchmark in the field. The code will be available at https://github.com/guobaoxiao/SAM-TTT.

Problem

• Enhancing camouflaged object detection via SAM parameter optimization
• Addressing adverse parameter effects on semantic understanding
• Integrating test-time training for vision task performance

Innovation

• Reverse SAM Parameter Configuration Module
• T-Visioner Module with Test-Time Training
• Simultaneously suppresses adverse and reinforces advantageous parameters