Modern Convolutional Neural Networks (CNNs) are usually restricted by their massive computation and high storage. Parameter pruning is a promising approach for CNN compression and acceleration by eliminating redundant model parameters with tolerable performance degradation. Despite its effectiveness, existing regularization-based parameter pruning methods usually drive weights towards zero with $large\ and\ constant$ regularization factors, which neglects the fragility of the expressiveness of CNNs, and thus calls for a more gentle regularization scheme so that the networks can adapt during pruning. To achieve this, we propose a novel regularization-based pruning method, named $IncReg$ , to $incrementally$ assign different regularization factors to different weights based on their relative importance. Empirical analysis on CIFAR-10 dataset verifies the merits of IncReg. Further extensive experiments with popular CNNs on CIFAR-10 and ImageNet datasets show that IncReg achieves comparable to even better results compared with state-of-the-arts. Moreover, to resolve the problem that column pruning cannot be directly applied to off-the-shelf deep learning libraries for acceleration, we generalize IncReg from column pruning to spatial pruning, which can equip existing structured pruning methods (such as channel pruning) for further acceleration with ignorable accuracy loss. Our source codes and trained models are available here: https://github.com/mingsun-tse/caffe\_increg.

We demonstrate a silicon $1\times 64$ optical phased array by using carrier-depletion-based phase shifters. The 5-mm-long phase shifter exhibits an electro-optic bandwidth of 700 MHz, a modulation efficiency of 2.3 V.cm, and an ultra-low power consumption of < 0.45 nW/ $\boldsymbol {\pi }$ . A steering range of ±18° is realized in the phased-array axis by tuning the phase of each antenna. Besides, the beam steering in the other (wavelength-controlled) axis by tuning the wavelength achieves a dispersion efficiency of 0.17° /nm. Although the carrier-depletion effect induces a nonuniform optical amplitude distribution among the antenna array, the OPA chip is still able to form high-quality far-field beam patterns with side lobe suppression ratios of $\boldsymbol {\sim }10$ dB. Our measurement and simulation results indicate that the carrier-depletion-based OPA is a promising candidate for future high-speed and power-efficient solid-state Lidars.

Based on cascaded Mach-Zehnder interferometer (MZI) lattice filters, we demonstrate and compare silicon O-band 8-channel (de-)multiplexers with flat and Gaussian-like passbands for the 400 GBASE-LR8 norm. The 1×8 (de-)multiplexer with flat passbands exhibit insertion loss less than 2.5 dB, channel crosstalk lower than −11 dB, and 3 dB bandwidths of ∼3.9 nm. In contrast, the maximum insertion loss, the worst channel crosstalk, and 3 dB bandwidths of the 1×8 (de-)multiplexer with Gaussian-like passbands are 1.1 dB, −13 dB, and ∼2.9 nm, respectively. The comparison result indicates that the latter is more advisable choice. A theoretical calculation based on measured coupling ratios of stand-along directional couplers (DC) indicates that the performance deteriorations of practical filtering curves are mainly caused by dispersions and fabrication errors of the DCs.

A recommendation system is often used to recommend items that may be of interest to users. One of the main challenges is that the scarcity of actual interaction data between users and items restricts the performance of recommendation systems. To solve this problem, multi-modal technologies have been used for expanding available information. However, the existing multi-modal recommendation algorithms all extract the feature of single modality and simply splice the features of different modalities to predict the recommendation results. This fusion method can not completely mine the relevance of multi-modal features and lose the relationship between different modalities, which affects the prediction results. In this paper, we propose a Cross-Modal-Based Fusion Recommendation Algorithm (CMBF) that can capture both the single-modal features and the cross-modal features. Our algorithm uses a novel cross-modal fusion method to fuse the multi-modal features completely and learn the cross information between different modalities. We evaluate our algorithm on two datasets, MovieLens and Amazon. Experiments show that our method has achieved the best performance compared to other recommendation algorithms. We also design ablation study to prove that our cross-modal fusion method improves the prediction results.

We demonstrate an optical switch array based on the topology of Butterfly by using dual-microring resonators which can be used in the ultra-compact network. The proposed switch is composed of 12 basic switch elements tuned by thermo-optic (TO) effect, which can build links freely without any blocking. Due to the advantages of microring resonators, the footprint switch array is smaller than 0.2 mm ² . The optical structure can complete 4- and 8-ports signal routing. It shows a large bandwidth over 50GHz, thereby performing well under condition of high-speed data transmission.

The performance of voice-controlled systems is usually influenced by accented speech. To make these systems more robust, frontend accent recognition (AR) technologies have received increased attention in recent years. As accent is a high-level abstract feature that has a profound relationship with language knowledge, AR is more challenging than other language-agnostic audio classification tasks. In this paper, we use an auxiliary automatic speech recognition (ASR) task to extract language-related phonetic features. Furthermore, we propose a hybrid structure that incorporates the embeddings of both a fixed acoustic model and a trainable acoustic model, making the language-related acoustic feature more robust. We conduct several experiments on the AESRC dataset. The results demonstrate that our approach can obtain an 8.02% relative improvement compared with the Transformer baseline, showing the merits of the proposed method.