--Visual Place Recognition (VPR) is crucial for robust mobile robot localization, yet it faces significant challenges in maintaining reliable performance under varying environmental conditions and viewpoints. T o address this, we propose a novel framework that integrates Dual-Scale-Former (DSFormer), a Transformer-based cross-learning module, with an innovative block clustering strategy. DSFormer enhances feature representation by enabling bidirectional information transfer between dual-scale features extracted from the final two CNN layers, capturing both semantic richness and spatial details through self-attention for long-range dependencies within each scale and shared cross-attention for cross-scale learning. Complementing this, our block clustering strategy repartitions the widely used San Francisco eXtra Large (SF-XL) training dataset from multiple distinct perspectives, optimizing data organization to further bolster robustness against viewpoint variations. T ogether, these innovations not only yield a robust global embedding adaptable to environmental changes but also reduce the required training data volume by approximately 30% compared to previous partitioning methods. Comprehensive experiments demonstrate that our approach achieves state-of-the-art performance across most benchmark datasets, surpassing advanced reranking methods like DELG, Patch-NetVLAD, TransVPR, and R2Former as a global retrieval solution using 512-dim global descriptors, while significantly improving computational efficiency. PR serves as a fundamental capability in robotic systems, enabling robots to coarsely locate themselves within an environment by matching visual inputs to a pre-existing geo-tagged database, which is critical for robotics to large-scale geolocation tasks.

In panoptic scene graph generation (PSGG), models retrieve interactions between objects in an image which are grounded by panoptic segmentation masks. Previous evaluations on panoptic scene graphs have been subject to an erroneous evaluation protocol where multiple masks for the same object can lead to multiple relation distributions per mask-mask pair. This can be exploited to increase the final score. We correct this flaw and provide a fair ranking over a wide range of existing PSGG models. The observed scores for existing methods increase by up to 7.4 mR@50 for all two-stage methods, while dropping by up to 19.3 mR@50 for all one-stage methods, highlighting the importance of a correct evaluation. Contrary to recent publications, we show that existing two-stage methods are competitive to one-stage methods. Building on this, we introduce the Decoupled SceneFormer (DSFormer), a novel two-stage model that outperforms all existing scene graph models by a large margin of +11 mR@50 and +10 mNgR@50 on the corrected evaluation, thus setting a new SOTA. As a core design principle, DSFormer encodes subject and object masks directly into feature space.

With the tremendous success of large transformer models in natural language understanding, down-sizing them for cost-effective deployments has become critical. Recent studies have explored the low-rank weight factorization techniques which are efficient to train, and apply out-of-the-box to any transformer architecture. Unfortunately, the low-rank assumption tends to be over-restrictive and hinders the expressiveness of the compressed model. This paper proposes, DSFormer, a simple alternative factorization scheme which expresses a target weight matrix as the product of a small dense and a semi-structured sparse matrix. The resulting approximation is more faithful to the weight distribution in transformers and therefore achieves a stronger efficiency-accuracy trade-off. Another concern with existing factorizers is their dependence on a task-unaware initialization step which degrades the accuracy of the resulting model. DSFormer addresses this issue through a novel Straight-Through Factorizer (STF) algorithm that jointly learns all the weight factorizations to directly maximize the final task accuracy. Extensive experiments on multiple natural language understanding benchmarks demonstrate that DSFormer obtains up to 40% better compression than the state-of-the-art low-rank factorizers, leading semi-structured sparsity baselines and popular knowledge distillation approaches. Our approach is also orthogonal to mainstream compressors and offers up to 50% additional compression when added to popular distilled, layer-shared and quantized transformers. We empirically evaluate the benefits of STF over conventional optimization practices.

Multivariate time series long-term prediction, which aims to predict the change of data in a long time, can provide references for decision-making. Although transformer-based models have made progress in this field, they usually do not make full use of three features of multivariate time series: global information, local information, and variables correlation. To effectively mine the above three features and establish a high-precision prediction model, we propose a double sampling transformer (DSformer), which consists of the double sampling (DS) block and the temporal variable attention (TVA) block. Firstly, the DS block employs down sampling and piecewise sampling to transform the original series into feature vectors that focus on global information and local information respectively. Then, TVA block uses temporal attention and variable attention to mine these feature vectors from different dimensions and extract key information. Finally, based on a parallel structure, DSformer uses multiple TVA blocks to mine and integrate different features obtained from DS blocks respectively. The integrated feature information is passed to the generative decoder based on a multi-layer perceptron to realize multivariate time series long-term prediction. Experimental results on nine real-world datasets show that DSformer can outperform eight existing baselines.

Multi-contrast MRI (MC-MRI) captures multiple complementary imaging modalities to aid in radiological decision-making. Given the need for lowering the time cost of multiple acquisitions, current deep accelerated MRI reconstruction networks focus on exploiting the redundancy between multiple contrasts. However, existing works are largely supervised with paired data and/or prohibitively expensive fully-sampled MRI sequences. Further, reconstruction networks typically rely on convolutional architectures which are limited in their capacity to model long-range interactions and may lead to suboptimal recovery of fine anatomical detail. To these ends, we present a dual-domain self-supervised transformer (DSFormer) for accelerated MC-MRI reconstruction. DSFormer develops a deep conditional cascade transformer (DCCT) consisting of several cascaded Swin transformer reconstruction networks (SwinRN) trained under two deep conditioning strategies to enable MC-MRI information sharing. We further present a dual-domain (image and k-space) self-supervised learning strategy for DCCT to alleviate the costs of acquiring fully sampled training data. DSFormer generates high-fidelity reconstructions which experimentally outperform current fully-supervised baselines. Moreover, we find that DSFormer achieves nearly the same performance when trained either with full supervision or with our proposed dual-domain self-supervision.