Smiling has often been incorrectly interpreted as “happy” in the popular facial expression datasets (AffectNet, RAF-DB, FERPlus). Smiling is the most complex human expression, with positive, neutral, and negative smiles. We focused on fine-grained facial expression recognition (FER) and built a new smiling face dataset, named Facial Expression Emotions. This dataset categorizes smiles into six classes of smiles, containing a total of 11,000 images labeled with corresponding fine-grained facial expression classes. We propose Smile Transformer, a network architecture for FER based on the Swin Transformer, to enhance the local perception capability of the model and improve the accuracy of fine-grained face recognition. Moreover, a convolutional block attention module (CBAM) was designed, to focus on important features of the face image and suppress unnecessary regional responses. For better classification results, an image quality evaluation module was used to assign different labels to images with different qualities. Additionally, a dynamic weight loss function was designed, to assign different learning strategies according to the labels during training, focusing on hard yet recognizable samples and discarding unidentifiable samples, to achieve better recognition. Overall, we focused on (a) creating a novel dataset of smiling facial images from online annotated images, and (b) developing a method for improved FER in smiling images. Facial Expression Emotions achieved an accuracy of 88.56% and could serve as a new benchmark dataset for future research on fine-grained FER.

Although the NeRF approach can achieve outstanding view synthesis, it is limited in practical use because it requires many views (hundreds) for training. With only a few input views, the Depth-DYN NeRF that we propose can accurately match the shape. First, we adopted the ip_basic depth-completion method, which can recover the complete depth map from sparse radar depth data. Then, we further designed the Depth-DYN MLP network architecture, which uses a dense depth prior to constraining the NeRF optimization and combines the depthloss to supervise the Depth-DYN MLP network. When compared to the color-only supervised-based NeRF, the Depth-DYN MLP network can better recover the geometric structure of the model and reduce the appearance of shadows. To further ensure that the depth depicted along the rays intersecting these 3D points is close to the measured depth, we dynamically modified the sample space based on the depth of each pixel point. Depth-DYN NeRF considerably outperforms depth NeRF and other sparse view versions when there are a few input views. Using only 10–20 photos to render high-quality images on the new view, our strategy was tested and confirmed on a variety of benchmark datasets. Compared with NeRF, we obtained better image quality (NeRF average at 22.47 dB vs. our 27.296 dB).

Neural radiation field (NeRF)-based novel view synthesis methods are gaining popularity. NeRF can generate more detailed and realistic images than traditional methods. Conventional NeRF reconstruction of a room scene requires at least several hundred images as input data and generates several spatial sampling points, placing a tremendous burden on the training and prediction process with respect to memory and computational time. To address these problems, we propose a prior-driven NeRF model that only accepts sparse views as input data and reduces a significant number of non-functional sampling points to improve training and prediction efficiency and achieve fast high-quality rendering. First, this study uses depth priors to guide sampling, and only a few sampling points near the controllable range of the depth prior are used as input data, which reduces the memory occupation and improves the efficiency of training and prediction. Second, this study encodes depth priors as distance weights into the model and guides the model to quickly fit the object surface. Finally, a novel approach combining the traditional mesh rendering method (TMRM) and the NeRF volume rendering method was used to further improve the rendering efficiency. Experimental results demonstrated that our method had significant advantages in the case of sparse input views (11 per room) and few sampling points (8 points per ray).