On the robustness of skeleton detection against adversarial attacks

CONet: Crowd and occlusion-aware network for occluded human pose estimation

Human pose estimation has numerous applications in motion recognition, virtual reality, human-computer interaction, and other related fields. However, multi-person pose estimation in crowded and occluded scenes is challenging. One major issue about the current top-down human pose estimation approaches is that they are limited to predicting the pose of a single person, even when the bounding box contains multiple individuals. To address this problem, we propose a novel Crowd and Occlusion-aware Network (CONet) using a divide-and-conquer strategy. Our approach includes a Crowd and Occlusion-aware Head (COHead) which estimates the pose of both the occluder and the occluded person using two separate branches. We also use the attention mechanism to guide the branches for differentiated learning, aiming to improve feature representation. Additionally, we propose a novel interference point loss to enhance the model's anti-interference ability. Our CONet is simple yet effective, and it outperforms the state-of-the-art model by +1.6 AP, achieving 71.6 AP on CrowdPose. Our proposed model has achieved state-of-the-art results on the CrowdPose dataset, demonstrating its effectiveness in improving the accuracy of human pose estimation in crowded and occluded scenes. This achievement highlights the potential of our model in many real-world applications where accurate human pose estimation is crucial, such as surveillance, sports analysis, and human-computer interaction.

Are transformer-based models more robust than CNN-based models?

As the deployment of artificial intelligence (AI) models in real-world settings grows, their open-environment robustness becomes increasingly critical. This study aims to dissect the robustness of deep learning models, particularly comparing transformer-based models against CNN-based models. We focus on unraveling the sources of robustness from two key perspectives: structural and process robustness. Our findings suggest that transformer-based models generally outperform convolution-based models in robustness across multiple metrics. However, we contend that these metrics may not wholly represent true model robustness, such as the mean of corruption error. To better understand the underpinnings of this robustness advantage, we analyze models through the lens of Fourier transform and game interaction. From our insights, we propose a calibrated evaluation metric for robustness against real-world data, and a blur-based method to enhance robustness performance. Our approach achieves state-of-the-art results, with mCE scores of 2.1% on CIFAR-10-C, 12.4% on CIFAR-100-C, and 24.9% on TinyImageNet-C.

Development and Validation of a Deep Learning Model to Reduce the Interference of Rectal Artifacts in MRI-based Prostate Cancer Diagnosis

Purpose To develop an MRI-based model for clinically significant prostate cancer (csPCa) diagnosis that can resist rectal artifact interference. Materials and Methods This retrospective study included 2203 male patients with prostate lesions who underwent biparametric MRI and biopsy between January 2019 and June 2023. Targeted adversarial training with proprietary adversarial samples (TPAS) strategy was proposed to enhance model resistance against rectal artifacts. The automated csPCa diagnostic models trained with and without TPAS were compared using multicenter validation datasets. The impact of rectal artifacts on the diagnostic performance of each model at the patient and lesion levels was compared using the area under the receiver operating characteristic curve (AUC) and the area under the precision-recall curve (AUPRC). The AUC between models was compared using the DeLong test, and the AUPRC was compared using the bootstrap method. Results The TPAS model exhibited diagnostic performance improvements of 6% at the patient level (AUC: 0.87 vs 0.81, P < .001) and 7% at the lesion level (AUPRC: 0.84 vs 0.77, P = .007) compared with the control model. The TPAS model demonstrated less performance decline in the presence of rectal artifact-pattern adversarial noise than the control model (ΔAUC: -17% vs -19%, ΔAUPRC: -18% vs -21%). The TPAS model performed better than the control model in patients with moderate (AUC: 0.79 vs 0.73, AUPRC: 0.68 vs 0.61) and severe (AUC: 0.75 vs 0.57, AUPRC: 0.69 vs 0.59) artifacts. Conclusion This study demonstrates that the TPAS model can reduce rectal artifact interference in MRI-based csPCa diagnosis, thereby improving its performance in clinical applications. Keywords: MR-Diffusion-weighted Imaging, Urinary, Prostate, Comparative Studies, Diagnosis, Transfer Learning Clinical trial registration no. ChiCTR23000069832 Supplemental material is available for this article. Published under a CC BY 4.0 license.

Strengthening transferability of adversarial examples by adaptive inertia and amplitude spectrum dropout

Deep neural networks are sensitive to adversarial examples and would produce wrong results with high confidence. However, most existing attack methods exhibit weak transferability, especially for adversarially trained models and defense models. In this paper, two methods are proposed to generate highly transferable adversarial examples, namely Adaptive Inertia Iterative Fast Gradient Sign Method (AdaI2-FGSM) and Amplitude Spectrum Dropout Method (ASDM). Specifically, AdaI2-FGSM aims to integrate adaptive inertia into the gradient-based attack, and leverage the looking ahead property to search for a flatter maximum, which is essential to strengthen the transferability of adversarial examples. By introducing a loss-preserving transformation in the frequency domain, the proposed ASDM with the dropout invariance property can craft the copies of input images to overcome the poor generalization on the surrogate models. Furthermore, AdaI2-FGSM and ASDM can be naturally integrated as an efficient gradient-based attack method to yield more transferable adversarial examples. Extensive experimental results on the ImageNet-compatible dataset demonstrate that higher transferability is achieved by our method than some advanced gradient-based attacks.

ProMask: Probability mask representation for skeleton detection

Detecting object skeletons in natural images presents challenges due to varied object scales and complex backgrounds. The skeleton is a highly compressing shape representation, which can bring some essential advantages but cause difficulties in detection. This skeleton line occupies a small part of the image and is overly sensitive to spatial position. Inspired by these issues, we propose the ProMask, which is a novel skeleton detection model. The ProMask includes the probability mask representation and vector router. This skeleton probability mask describes the gradual formation process of skeleton points, which can achieve high detection performance and robustness. Moreover, the vector router module possesses two sets of orthogonal basis vectors in a two-dimensional space, which can dynamically adjust the predicted skeleton position. Experiments show that our approach realizes better performance, efficiency, and robustness than state-of-the-art methods. We consider that our proposed skeleton probability representation will serve as a standard configuration for future skeleton detection, since it is reasonable, simple, and very effective.

Deep learning approach for guiding three-dimensional computed tomography reconstruction of lower limbs for robotically-assisted total knee arthroplasty

BACKGROUND

Robotic-assisted total knee arthroplasty (TKA) was performed to promote the accuracy of bone resection and mechanical alignment. Among these TKA system procedures, 3D reconstruction of CT data of lower limbs consumes significant manpower. Artificial intelligence (AI) algorithms applying deep learning has been proved efficient in automated identification and visual processing.

METHODS

CT data of a total of 200 lower limbs scanning were used for AI-based 3D model construction and CT data of 20 lower limbs scanning were utilised for verification.

RESULTS

We showed that the performance of an AI-guided 3D reconstruction of CT data of lower limbs for robotic-assisted TKA was similar to that of the operator-based approach. The time of 3D lower limb model construction using AI was 4.7 min. AI-based 3D models can be used for surgical planning.

CONCLUSION

AI was used for the first time to guide the 3D reconstruction of CT data of lower limbs for facilitating robotic-assisted TKA. Incorporation of AI in 3D model reconstruction before TKA might reduce the workload of radiologists.