SAM2Dual: Training-Free, Dual Memory for Long-Term Video Object Segmentation

• URL: http://arxiv.org/abs/2608.18640v1
• Date: Wed, 19 Aug 2026 07:40:39 GMT
• Title: SAM2Dual: Training-Free, Dual Memory for Long-Term Video Object Segmentation
• Abstract summary: We propose SAM2Dual, a training-free, plug-and-play inference-time enhancement that improves long-video robustness without updating model weights.<n> SAM2Dual introduces a Dual Memory design that explicitly separates (i) short-term memory for rapid local adaptation and (ii) long-term memory built via interval-based sampling to preserve global identity cues.<n>In addition, we present Text-Aware Memory (TAM), which extracts a compact word-level cue from early frames and uses text embeddings to reweight memory contributions based on semantic compatibility.
• Score: 2.3605238326156806
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Long-term video object segmentation (VOS) remains challenging due to error accumulation under extended occlusions, re-appearance, and scene changes. Although SAM2 provides strong zero-shot performance, its streaming memory can amplify drift over long horizons when recent, unreliable predictions dominate the memory state. We propose SAM2Dual, a training-free, plug-and-play inference-time enhancement that improves long-video robustness without updating model weights. SAM2Dual introduces a Dual Memory design that explicitly separates (i) short-term memory for rapid local adaptation and (ii) long-term memory built via interval-based sampling to preserve global identity cues, combined through a gated fusion strategy. In addition, we present Text-Aware Memory (TAM), which extracts a compact word-level cue from early frames and uses text embeddings to reweight memory contributions based on semantic compatibility, supporting identity preservation when visual evidence becomes weak or ambiguous. Across long-term benchmarks, SAM2Dual consistently improves stability on long videos, raising J&F from 49.33 to 50.65 on MOSEv2 and achieving consistent gains on LVOSv2.

Related papers

• Semantic-Aware Adaptive Visual Memory for Streaming Video Understanding [55.7992006853979]
  SAVEMem is a framework that brings semantic awareness into memory generation and lets the retrieval scope adapt per query.<n>We present SAVEMem, a training-free dual-stage framework that brings semantic awareness into memory generation and lets the retrieval scope adapt per query.
  arXiv  Detail & Related papers  (2026-05-08T15:40:40Z)
• Robust Ego-Exo Correspondence with Long-Term Memory [34.992180181705]
  We present a novel framework for establishing object-level correspondence between egocentric and exocentric views.<n>Our approach features a dual-memory architecture and an adaptive feature routing module inspired by Mixture-of-Experts (MoE)<n>In the experiments on the challenging EgoExo4D benchmark, our method, dubbed LM-EEC, achieves new state-of-the-art results.
  arXiv  Detail & Related papers  (2025-10-13T13:54:12Z)
• SAM2RL: Towards Reinforcement Learning Memory Control in Segment Anything Model 2 [2.659882635924329]
  Segment Anything Model 2 (SAM 2) has demonstrated strong performance in object segmentation tasks.<n>Recent methods augment SAM 2 with hand-crafted update rules to better handle distractors and object motion.<n>We propose a fundamentally different approach using reinforcement learning for optimizing memory updates in SAM 2.
  arXiv  Detail & Related papers  (2025-07-11T12:53:19Z)
• EdgeTAM: On-Device Track Anything Model [65.10032957471824]
  Segment Anything Model (SAM) 2 further extends its capability from image to video inputs through a memory bank mechanism.<n>We aim at making SAM 2 much more efficient so that it even runs on mobile devices while maintaining a comparable performance.<n>We propose EdgeTAM, which leverages a novel 2D Spatial Perceiver to reduce the computational cost.
  arXiv  Detail & Related papers  (2025-01-13T12:11:07Z)
• Look Every Frame All at Once: Video-Ma$^2$mba for Efficient Long-form Video Understanding with Multi-Axis Gradient Checkpointing [52.050036778325094]
  Video-Ma$2$mba is a novel architecture that incorporates State Space Models (SSMs) within the Mamba-2 framework.<n>Our approach significantly reduces the memory footprint compared to standard gradient checkpointing.<n>By maintaining a detailed capture of temporal dynamics, our model improves the accuracy and relevance of responses in long video understanding tasks.
  arXiv  Detail & Related papers  (2024-11-29T04:12:13Z)
• A Distractor-Aware Memory for Visual Object Tracking with SAM2 [11.864619292028278]
  Memory-based trackers are video object segmentation methods that form the target model by concatenating recently tracked frames into a memory buffer and localize the target by attending the current image to the buffered frames.<n> SAM2.1++ outperforms SAM2.1 and related SAM memory extensions on seven benchmarks and sets a solid new state-of-the-art on six of them.
  arXiv  Detail & Related papers  (2024-11-26T16:41:09Z)
• Stable Hadamard Memory: Revitalizing Memory-Augmented Agents for Reinforcement Learning [64.93848182403116]
  Current deep-learning memory models struggle in reinforcement learning environments that are partially observable and long-term. We introduce the Stable Hadamard Memory, a novel memory model for reinforcement learning agents. Our approach significantly outperforms state-of-the-art memory-based methods on challenging partially observable benchmarks.
  arXiv  Detail & Related papers  (2024-10-14T03:50:17Z)
• Efficient Video Object Segmentation via Modulated Cross-Attention Memory [123.12273176475863]
  We propose a transformer-based approach, named MAVOS, to model temporal smoothness without requiring frequent memory expansion. Our MAVOS achieves a J&F score of 63.3% while operating at 37 frames per second (FPS) on a single V100 GPU.
  arXiv  Detail & Related papers  (2024-03-26T17:59:58Z)
• Recurrent Dynamic Embedding for Video Object Segmentation [54.52527157232795]
  We propose a Recurrent Dynamic Embedding (RDE) to build a memory bank of constant size. We propose an unbiased guidance loss during the training stage, which makes SAM more robust in long videos. We also design a novel self-correction strategy so that the network can repair the embeddings of masks with different qualities in the memory bank.
  arXiv  Detail & Related papers  (2022-05-08T02:24:43Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.