RESUMEN

Molecular dynamics simulations in the microcanonical ensemble are performed to study the collapse of a bubble in liquid water using the single-site mW and the four-site TIP4P/2005 water models. To study system size effects, simulations for pure water systems are performed using periodically replicated simulation boxes with linear dimensions, L, ranging from 32 to 512 nm with the largest systems containing 8.7 × 106 and 4.5 × 109 molecules for the TIP4P/2005 and mW water models, respectively. The computationally more efficient mW water model allows us to reach converging behavior when the bubble dynamics results are plotted in reduced units, and the limiting behavior can be obtained through linear extrapolation in L-1. Qualitative differences are observed between simulations with the mW and TIP4P/2005 water models, but they can be explained by the models' differences in predicted viscosity and surface tension. Although bubble collapse occurs on time scales of only hundreds of picoseconds, the system sizes used here are sufficiently large to obtain bubble dynamics consistent with the Rayleigh-Plesset equation when using the models' thermophysical properties as input. For the conditions explored here, extreme heating of the interfacial water molecules near the time of collapse is observed for the larger mW water systems (but the model underpredicts the viscosity), whereas heating is less pronounced for the TIP4P/2005 water systems because its larger viscosity contribution slows the collapse dynamics. The presence of nitrogen within the bubble only starts to affect bubble dynamics near the very end of the initial collapse, leading to an incomplete collapse and strong rebound for the mW water model. Although nitrogen is non-condensable at 300 K, it becomes highly compressed and reaches a liquid-like density near the collapse point. We find that the dissolution of nitrogen is much slower than the movement of the collapsing water front, and the re-expansion of the dense nitrogen droplet gives rise to bubble rebound. The incompatibility of the collapse and dissolution time scales should be considered for continuum-scale modeling of bubble dynamics. We also confirm that the diffusion coefficient for dissolved nitrogen is insensitive to pressure as the liquid transitions from a compressed to a stretched state.

RESUMEN

The widespread dissemination of facial forgery technology has brought many ethical issues and aroused widespread concern in society. Most research today treats deepfake detection as a fine grained classification task, which however makes it difficult to enable the feature extractor to express the features related to the real and fake attributes. This paper proposes a depth map guided triplet network, which mainly consists of a depth prediction network and a triplet feature extraction network. The depth map predicted by the depth prediction network can effectively reflect the differences between real and fake faces in discontinuity, inconsistent illumination, and blurring, thus in favor of deepfake detection. Regardless of the facial appearance changes induced by deepfake, we argue that real and fake faces should correspond to their respective latent feature spaces. Particularly, the pair of real faces (original-target) remain close in the latent feature space, while the two pairs of real-fake faces (original-fake, target-fake) instead keep faraway. Following this paradigm, we suggest a triplet loss supervision network to extract the sufficiently discriminative deep features, which minimizes the distance of the original-target pair and maximize the distance of the original-fake (also target-fake) pair. The extensive results on public FaceForensics++ and Celeb-DF datasets validate the superiority of our method over competitors.

Asunto(s)

RESUMEN

Because the data volume of news videos is increasing exponentially, a way to quickly browse a sketch of the video is important in various applications, such as news media, archives and publicity. This paper proposes a news video summarization method based on SURF features and an improved clustering algorithm, to overcome the defects in existing algorithms that fail to account for changes in shot complexity. Firstly, we extracted SURF features from the video sequences and matched the features between adjacent frames, and then detected the abrupt and gradual boundaries of the shot by calculating similarity scores between adjacent frames with the help of double thresholds. Secondly, we used an improved clustering algorithm to cluster the color histogram of the video frames within the shot, which merged the smaller clusters and then selected the frame closest to the cluster center as the key frame. The experimental results on both the public and self-built datasets show the superiority of our method over the alternatives in terms of accuracy and speed. Additionally, the extracted key frames demonstrate low redundancy and can credibly represent a sketch of news videos.