As an intersection of artificial intelligence and human communication analysis, Emotion Recognition in Conversation (ERC) has attracted much research attention in recent years. Existing studies, however, are limited in adequately exploiting latent relations among the constituent utterances. In this paper, we address this issue by proposing a novel approach named Latent Relation-Aware Graph Convolutional Network (LR-GCN), where both speaker dependency of the interlocutors is leveraged and latent correlations among the utterances are captured for ERC. Specifically, we first establish a graph model to incorporate the context information and speaker dependency of the conversation. Afterward, the multi-head attention mechanism is introduced to explore the latent correlations among the utterances and generate a set of all-linked graphs. Here, aiming to simultaneously exploit the original modeled speaker dependency and the explored correlation information, we introduce a dense connection layer to capture more structural information of the generated graphs. Through a multi-branch graph network, we achieve a unified representation of each utterance for final prediction. Detailed evaluations on two benchmark datasets demonstrate LR-GCN outperforms the state-of-the-art approaches.

Developing an intrusion event identification scheme, which concurrently possesses high recognition accuracy and high efficiency, has been an intractable task in optical fiber perimeter security systems. To achieve high recognition accuracy, we apply the all-phase filter (APF) bank in frequency division and choose the envelope fiuctuation parameter to describe the waveform feature of APFs' outputs. To achieve high recognition efficiency, we introduce the random forest classifier to recognize intrusion types, which not only alleviates the negative effect arising from occasionality or randomness of intrusions, but also bypasses tedious computation of existing classifiers applied to optical fiber dual Mach-Zehnder Interferometry based perimeter security system. Experimental results demonstrate that the proposed system can distinguish 6 typical patterns (kicking the fence, cuttingthefence,wagglingthefence,knockingthefence,climbing the fence, and no intrusion) with the average recognition rate of 96.92%. Moreover, the consumed training time is reduced to about 40% of the Support Vector Machine. Therefore, the proposed scheme has vast potentials in actual applications.

Multimodal emotion recognition in conversations (ERC) aims to identify the emotional state of constituent utterances expressed by multiple speakers in dialogue from multimodal data. Existing multimodal ERC approaches focus on modeling the global context of the dialogue and neglect to mine the characteristic information from the corresponding utterances expressed by the same speaker. Additionally, information from different modalities exhibits commonality and diversity for emotional expression. The commonality and diversity of multimodal information are compensated for each other but not effectively exploited in previous multimodal ERC works. To tackle these issues, we propose a novel Multimodal Adversarial Learning Network (MALN). MALN first mines the speaker’s characteristics from context sequences and then incorporate them with the unimodal features. Afterward, we design a novel adversarial module AMDM to exploit both commonality and diversity from the unimodal features. Finally, AMDM fuses different modalities to generate refined utterance representations for emotion classification. Extensive experiments are conducted on two public multimodal ERC datasets, IEMOCAP and MELD. Through the experiments, MALN shows its superiority over the state-of-the-art methods.

A dual-function solar cell antenna is presented for wideband communication and optical energy harvesting. In the proposed design, four solar cells have been adopted as antenna and energy harvesting array for both wireless communication and optical energy harvesting simultaneously. The radiation performances of the proposed solar cell antenna are measured and the optical experiment is conducted to verify its abilities of optical energy harvesting. To clarify the effects of solar cells on the antenna performance, another copper antenna with the same structure is fabricated for comparison. The proposed technology can integrate wireless communication and green battery components, makes it promising for green communication, clean energy, and environmental protection.

The non-contact detection of human vital signs (i.e., respiration rate (RR) and heartbeat rate (HR)) using a continuous-wave (CW) Doppler radar sensor has great potential for intensive care monitoring, home healthcare, etc. However, large-scale and fast random body movement (RBM) has been a bottleneck for vital sign detection using a single CW Doppler radar. To break this dilemma, this study proposed a scheme combining adaptive noise cancellation (ANC) with polynomial fitting, which could retrieve the weak components of both respiration and heartbeat signals that were submerged under serious RBM interference. In addition, the new-type discrete cosine transform (N-DCT) was introduced to improve the detection accuracy. This scheme was first verified using a numerical simulation. Then, experiments utilizing a 10-GHz Doppler radar sensor that was built from general-purpose radio frequency (RF) and communication instruments were also carried out. No extra RF/microwave components and modules were needed, and neither was a printed circuit board nor an integrated-chip design required. The experimental results showed that both the RR and HR could still be extracted during large-scale and fast body movements using only a single Doppler radar sensor because the RBM noises could be greatly eliminated by utilizing the proposed ANC algorithm.

A low-profile and wideband patch antenna is proposed for base station applications. In order to realize a wide band, two types of slots are adopted and three resonant modes are constructed within a band of interest. Mode analysis are conducted to clarify the wideband performance. The designed antenna has an impedance bandwidth of 52.7% (from 1.62 GHz to 2.78 GHz) with a low-profile structure of 0.7$\lambda_{0}\times$ 0.47$\lambda_{0}\times$ 0.084$\lambda_{0}$, where $\lambda_{0}$ is the free space wavelength at the center frequency. The measurement results of a prototype agree well with the simulations. The measured average gain is 6 dBi and the radiation patterns remain stable in the required bands with low cross-polarization, low back lobes, nearly identical E- and H-plane patterns. The proposed multi-resonant antenna can cover six bands of wireless communication, such as DCS, PCS, UMTS, WLAN, LTE2300 and LTE2600, exhibiting great potentials in applications of multi-band wireless communication systems.