Weight bearing during exercise plays an important role in improving the mechanical properties of bone. The effect on bone of non-weight-bearing exercise such as swimming remains controversial. To investigate the effects of exercise mode on growing bone, 29 male Wistar rats (7 wk old) were randomly assigned to a running exercise group (Run, n = 9), a swimming exercise group (Swim, n = 10), or a nonexercise control group (Con, n = 10). During an 8-wk training session (20-60 min/day, 5 days/wk), the Run rats were trained at progressively increasing running speeds (12-22 m/min), and weights attached to the tail of the Swim rats were progressively increased from 0 to 2% of their body weight. The bone mineral density of the proximal tibiae of the Run rats was significantly higher than in the Swim (P < 0.05). Femoral wet weights of the two exercise groups were significantly higher than in the control group (P < 0.05). Interestingly, the percent difference between the tissue wet weight and dry weight (water content ratio), which is related to bone mechanical properties, was significantly higher in the tibiae of the Swim rats and the femora of both exercise groups compared with controls (P < 0.05). Extrinsic as well as intrinsic biomechanical material properties were measured in a three-point bending test. Bone mechanical properties of the tibiae and femora of rats in the Swim and Run groups were significantly greater than those in the control group (P < 0.05). In summary, different modes of exercise may benefit bone mechanical properties in different ways. The specific effects of swimming exercise (non-weight-bearing exercise) on bone require further study.

We show how a simple convolutional neural network (CNN) can be trained to accurately and robustly regress 6 degrees of freedom (6DoF) 3D head pose, directly from image intensities. We further explain how this FacePoseNet (FPN) can be used to align faces in 2D and 3D as an alternative to explicit facial landmark detection for these tasks. We claim that in many cases the standard means of measuring landmark detector accuracy can be misleading when comparing different face alignments. Instead, we compare our FPN with existing methods by evaluating how they affect face recognition accuracy on the IJB-A and IJB-B benchmarks: using the same recognition pipeline, but varying the face alignment method. Our results show that (a) better landmark detection accuracy measured on the 300W benchmark does not necessarily imply better face recognition accuracy. (b) Our FPN provides superior 2D and 3D face alignment on both benchmarks. Finally, (c), FPN aligns faces at a small fraction of the computational cost of comparably accurate landmark detectors. For many purposes, FPN is thus a far faster and far more accurate face alignment method than using facial landmark detectors.

We describe a deep learning based method for estimating 3D facial expression coefficients. Unlike previous work, our process does not relay on facial landmark detection methods as a proxy step. Recent methods have shown that a CNN can be trained to regress accurate and discriminative 3D morphable model (3DMM) representations, directly from image intensities. By foregoing facial landmark detection, these methods were able to estimate shapes for occluded faces appearing in unprecedented in-the-wild viewing conditions. We build on those methods by showing that facial expressions can also be estimated by a robust, deep, landmark-free approach. Our ExpNet CNN is applied directly to the intensities of a face image and regresses a 29D vector of 3D expression coefficients. We propose a unique method for collecting data to train this network, leveraging on the robustness of deep networks to training label noise. We further offer a novel means of evaluating the accuracy of estimated expression coefficients: by measuring how well they capture facial emotions on the CK+ and EmotiW-17 emotion recognition benchmarks. We show that our ExpNet produces expression coefficients which better discriminate between facial emotions than those obtained using state of the art, facial landmark detection techniques. Moreover, this advantage grows as image scales drop, demonstrating that our ExpNet is more robust to scale changes than landmark detection methods. Finally, at the same level of accuracy, our ExpNet is orders of magnitude faster than its alternatives.

This study investigated the effects of endurance running training on the bones of growing rats. Thirty-two male Wistar rats (7 weeks old) were assigned to a sedentary control group (CON, n = 10), a continuous endurance running group (CEN, n = 10), or an intermittent endurance running group (IEN, n = 12). After an 8-week training period, both exercise groups had significantly less body weight (BW) gain but higher aerobic capacity, shown by increased muscle citrate synthase (CS) activity. Bone area (BA), areal bone mineral density (aBMD), and bone mineral content (BMC) were measured by dual-energy Xray absorptiometry (DXA) in the total femur and sections of femora. Except for showing a significantly higher aBMD in total femora, the CON group was only slightly and nonsignificantly higher in other DXA measurements. In tissue weight measurements, the CON group showed a nonsignificantly higher tissue dry weight (P = 0.146), but a significantly lower tissue water content ratio (WCR, %) as compared to the exercise group. Despite having nonsignificantly lower long bone cross-sectional parameters, both exercise groups showed significantly better biomaterial properties, as measured by a three-point bending test. In extrinsic analysis, femora of the two exercise groups showed no difference in bending load and stiffness, but were significantly higher in post-yield bending energy and total ultimate bending energy (P < 0.05). Similar phenomena were revealed in tissue-level measurements; the CEN and IEN groups were significantly higher in ultimate toughness and post-yield toughness (P < 0.05). Higher post-yield energy shown by two exercise groups implied a change in bone matrix organization. In conclusion, this study demonstrated that two endurance treadmill training modes benefit bone, with subjects showing better tissue biomaterial properties without significantly increasing aBMD, BMC, or bone dimension. Further study would be valuable to investigate the effects of endurance running on other components of bone, such as organization of bone matrix and its relationship with bone biomaterial properties.