Monocular 3D Object Detection Based on Uncertainty Prediction of Keypoints

Three-dimensional (3D) object detection is an important task in the field of machine vision, in which the detection of 3D objects using monocular vision is even more challenging. We observe that most of the existing monocular methods focus on the design of the feature extraction framework or embedded geometric constraints, but ignore the possible errors in the intermediate process of the detection pipeline. These errors may be further amplified in the subsequent processes. After exploring the existing detection framework of keypoints, we find that the accuracy of keypoints prediction will seriously affect the solution of 3D object position. Therefore, we propose a novel keypoints uncertainty prediction network (KUP-Net) for monocular 3D object detection. In this work, we design an uncertainty prediction module to characterize the uncertainty that exists in keypoint prediction. Then, the uncertainty is used for joint optimization with object position. In addition, we adopt position-encoding to assist the uncertainty prediction, and use a timing coefficient to optimize the learning process. The experiments on our detector are conducted on the KITTI benchmark. For the two levels of easy and moderate, we achieve accuracy of 17.26 and 11.78 in AP3D, and achieve accuracy of 23.59 and 16.63 in APBEV, which are higher than the latest method KM3D.

Use of Multi-Seasonal Satellite Images to Predict SOC from Cultivated Lands in a Montane Ecosystem

Although algorithms are well developed to predict soil organic carbon (SOC), selecting appropriate covariates to improve prediction accuracy is an ongoing challenge. Terrain attributes and remote sensing data are the most common covariates for SOC prediction. This study tested the predictive performance of nine different combinations of topographic variables and multi-season remotely sensed data to improve the prediction of SOC in the cultivated lands of a middle mountain catchment of Nepal. The random forest method was used to predict SOC contents, and the quantile regression forest for quantifying the prediction uncertainty. Prediction of SOC contents was improved when remote sensing data of multiple seasons were used together with the terrain variables. Remote sensing data of multiple seasons capture the dynamic conditions of surface soils more effectively than using an image of a single season. It is concluded that the use of remote sensing images of multiple seasons instead of a snapshot of a single period may be more effective for improving the prediction of SOC in a digital soil mapping framework. However, an image with the right timing of cropping season can provide comparable results if a parsimonious model is preferred.

Face Swapping Consistency Transfer with Neural Identity Carrier

Deepfake aims to swap a face of an image with someone else’s likeness in a reasonable manner. Existing methods usually perform deepfake frame by frame, thus ignoring video consistency and producing incoherent results. To address such a problem, we propose a novel framework Neural Identity Carrier (NICe), which learns identity transformation from an arbitrary face-swapping proxy via a U-Net. By modeling the incoherence between frames as noise, NICe naturally suppresses its disturbance and preserves primary identity information. Concretely, NICe inputs the original frame and learns transformation supervised by swapped pseudo labels. As the temporal incoherence has an uncertain or stochastic pattern, NICe can filter out such outliers and well maintain the target content by uncertainty prediction. With the predicted temporally stable appearance, NICe enhances its details by constraining 3D geometry consistency, making NICe learn fine-grained facial structure across the poses. In this way, NICe guarantees the temporal stableness of deepfake approaches and predicts detailed results against over-smoothness. Extensive experiments on benchmarks demonstrate that NICe significantly improves the quality of existing deepfake methods on video-level. Besides, data generated by our methods can benefit video-level deepfake detection methods.

Uncertainty Prediction of Mining Safety Production Situation

In order to explore the occurrence and development law of mining safety production accidents, analyze its future change trends, and aim at the ambiguity, non-stationarity, and randomness of mining safety production accidents, an uncertainty prediction model for mining safety production situation is proposed. Firstly, the time series effect evaluation function is introduced to determine the optimal time granularity, which is used as the window width of fuzzy information granulation (FIG), and the time series of mining safety production situation is mapped to Low, R and Up three granular parameter sequences, according to the triangular fuzzy number; Then, the mean value of the intrinsic mode function (IMF) is maintained in the normal dynamic filtering range. After the ensemble empirical mode decomposition (EEMD), the three non-stationary granulation parameter sequences of Low, R and Up are decomposed into the intrinsic mode function components representing the detail information and the trend components representing the overall change, and then the sub-sequences are reconstructed according to the sample entropy to highlight the correlation among the sub-sequences; Finally, the cloud model language rules of mining safety production situation prediction are created. Through time series discretization, cloud transformation, concept jump, time series set division, association rule mining and uncertain reasoning, the reconstructed component sequence is modeled and predicted by uncertainty information extraction. The accuracy of the uncertainty prediction model was verified by 21 sets of test samples. The average relative errors of Low, R and Up sequences were 9.472 %, 16.671 % and 3.625 %, respectively. The research shows that the uncertainty prediction model of mining safety production situation overcomes the fuzziness, non-stationarity and uncertainty of safety production accidents, and provides theoretical reference and practical guidance for mining safety management and decision-making.

3D Primary Geochemical Halo Modeling and Its Application to the Ore Prediction of the Jiama Polymetallic Deposit, Tibet, China

The identification of primary geochemical haloes can be used to predict mineral resources in deep-seated orebodies through the delineation of element distributions. The Jiama deposits a typical skarn–porphyry Cu–polymetallic deposit in the Gangdese metallogenic belt of Tibet. The Cu–polymetallic skarn, Cu–Mo hornfels, and Mo ± Cu porphyry mineralization there exhibit superimposed geochemical haloes at depth. Three-dimensional (3D) primary geochemical halo modeling was undertaken for the deposit with the aim of providing geochemical data to describe element distributions in 3D space. An overall geochemical zonation of Zn(Pb) → Au → Cu(Ag) → Mo gained from geochemical cross-sections, together with dip-direction skarn zonation Pb–Zn(Cu) → Cu(Au–Ag–Mo) → Mo(Cu) → Cu–Mo(Au–Ag) and vertical zonation Cu–(Pb–Zn) → Mo–(Cu) → Mo–Cu–(Ag–Au–Pb–Zn) → Mo in the #24 exploration profile, indicates potential mineralization at depth. Integrated geochemical anomalies were extracted by kernel principal component analysis, which has the advantage of accommodating nonlinear data. A maximum-entropy model was constructed for deep mineral resources of uncertainty prediction. Three potential deep mineral targets are proposed on the basis of the obtained geochemical information and background.