Search for: All records

Image manipulation localization (IML) is a critical technique in media forensics, focusing on identifying tampered regions within manipulated images. Most existing IML methods require extensive training on labeled datasets with both image-level and pixel-level annotations. These methods often struggle with new manipulation types and exhibit low generalizability. In this work, we propose a training-free IML approach using diffusion models. Our method adaptively selects an appropriate number of diffusion timesteps for each input image in the forward process and performs both conditional and unconditional reconstructions in the backward process without relying on external conditions. By comparing these reconstructions, we generate a localization map highlighting regions of manipulation based on inconsistencies. Extensive experiments were conducted using sixteen state-of-the-art (SoTA) methods across six IML datasets. The results demonstrate that our training-free method outperforms SoTA unsupervised and weakly-supervised techniques. Furthermore, our method competes effectively against fully-supervised methods on novel (unseen) manipulation types. 

Abstract Physical processes behind flow‐topography interactions and turbulent transitions are essential for parameterization in numerical models. We examine how the Kuroshio cascades energy into turbulence upon passing over a seamount, employing a combination of shipboard measurements, tow‐yo microstructure profiling, and high‐resolution mooring. The seamount, spanning 5 km horizontally with two summits, interacts with the Kuroshio, whose flow speed ranges from 1 to 2 m s⁻¹, modulated by tides. The forward energy cascade process is commenced by forming a train of 2–3 nonlinear lee waves behind the summit with a wavelength of 0.5–1 km and an amplitude of 50–100 m. A train of Kelvin‐Helmholtz (KH) billows develops immediately below the lee waves and extends downstream, leading to enhanced turbulence. The turbulent kinetic energy dissipation rate isO(10⁻⁷–10⁻⁴) W kg⁻¹, varying in phase with the upstream flow speed modulated by tides. KH billows occur primarily at the lee wave's trailing edge, where the combined strong downstream shear and low‐stratification recirculation trigger the shear instability,Ri < 1/4. The recirculation also creates an overturn susceptible to gravitational instability. This scenario resembles the rotor, commonly found in atmospheric mountain waves but rarely observed in the ocean. A linear stability analysis further suggests that critical levels, where the KH instability extracts energy from the mean flow, are located predominantly at the strong shear layer of the lee wave's upwelling portion, coinciding with the upper boundary of the rotor. These novel observations may provide insights into flow‐topography interactions and improve physics‐based turbulence parameterization. 

Despite recent progress in Multiple Object Tracking (MOT), several obstacles such as occlusions, similar objects, and complex scenes remain an open challenge. Meanwhile, a systematic study of the cost-performance tradeoff for the popular tracking-by-detection paradigm is still lacking. This paper introduces SMILEtrack, an innovative object tracker that effectively addresses these challenges by integrating an efficient object detector with a Siamese network-based Similarity Learning Module (SLM). The technical contributions of SMILETrack are twofold. First, we propose an SLM that calculates the appearance similarity between two objects, overcoming the limitations of feature descriptors in Separate Detection and Embedding (SDE) models. The SLM incorporates a Patch Self-Attention (PSA) block inspired by the vision Transformer, which generates reliable features for accurate similarity matching. Second, we develop a Similarity Matching Cascade (SMC) module with a novel GATE function for robust object matching across consecutive video frames, further enhancing MOT performance. Together, these innovations help SMILETrack achieve an improved trade-off between the cost (e.g., running speed) and performance (e.g., tracking accuracy) over several existing state-of-the-art benchmarks, including the popular BYTETrack method. SMILETrack outperforms BYTETrack by 0.4-0.8 MOTA and 2.1-2.2 HOTA points on MOT17 and MOT20 datasets. Code is available at http://github.com/pingyang1117/SMILEtrack_official.