Detection of moving objects in a video captured by a freely moving camera is a challenging problem in computer vision. Most existing methods often assume that the background (BG) can be approximated by dominant single plane/multiple planes or impose significant geometric constraints on BG, or utilize a complex BG/foreground probabilistic model. Instead, we propose a computationally efficient algorithm that is able to detect moving objects accurately and robustly in a general 3D scene. This problem is formulated as a coarse-to-fine thresholding scheme on the particle trajectories in the video sequence. First, a coarse foreground (CFG) region is extracted by performing reduced singular value decomposition on multiple matrices that are built from bundles of particle trajectories. Next, the BG motion of pixels in the CFG region is reconstructed by a fast inpainting method. After subtracting the BG motion, the fine foreground is segmented out by an adaptive thresholding method that is capable of solving multiple-moving-objects scenarios. Finally, the detected foreground is further refined by the mean-shift segmentation method. Extensive simulations and a comparison with the state-of-the-art methods verify the effectiveness of the proposed method.

Microwaves have been widely used in the treatment of materials, such as heating, drying, and sterilization. However, the heating in the commonly used microwave applicators is usually uneven. In this paper, a novel multi-material turntable structure is creatively proposed to improve the temperature uniformity in microwave ovens. Three customized turntables consisting of polyethylene (PE) and alumina, PE and aluminum, and alumina and aluminum are, respectively, utilized in a domestic microwave oven in simulation. During the heating process, the processed material is placed on a fixed Teflon bracket which covers the constantly rotating turntable. Experiments are conducted to measure the surface and point temperatures using an infrared thermal imaging camera and optical fibers. Simulated results are compared qualitatively with the measured ones, which verifies the simulated models. Compared with the turntables consisting of a single material, a 26%–47% increase in temperature uniformity from adapting the multi-material turntable can be observed for the microwave-processed materials.

Referring expression comprehension (REC) is a vision-language task to locate a target object in an image based on a language expression. Fully fine-tuning general-purpose pre-trained vision-language foundation models for REC yields impressive performance but becomes increasingly costly. Parameter-efficient transfer learning (PETL) methods have shown strong performance with fewer tunable parameters. However, directly applying PETL to REC faces two challenges: (1) insufficient multi-modal interaction between pre-trained vision-language foundation models, and (2) high GPU memory usage due to gradients passing through the heavy vision-language foundation models. To this end, we present M2IST: Multi-Modal Interactive Side-Tuning with M3ISAs: Mixture of Multi-Modal Interactive Side-Adapters. During fine-tuning, we fix the pre-trained uni-modal encoders and update M3ISAs to enable efficient vision-language alignment for REC. Empirical results reveal that M2IST achieves better performance-efficiency trade-off than full fine-tuning and other PETL methods, requiring only 2.11% tunable parameters, 39.61% GPU memory, and 63.46% training time while maintaining competitive performance. Our code is released at https://github.com/xuyang-liu16/M2IST.