Aerial object detection plays a crucial role in numerous applications. However, most existing methods focus on detecting predefined object categories, limiting their applicability in real-world open scenarios. In this paper, we extend aerial object detection to open scenarios through image-text collaboration and propose RT-OVAD, the first real-time open-vocabulary detector for aerial scenes. Specifically, we first introduce an image-to-text alignment loss to replace the conventional category regression loss, thereby eliminating category constraints. Next, we propose a lightweight image-text collaboration strategy comprising an image-text collaboration encoder and a text-guided decoder. The encoder simultaneously enhances visual features and refines textual embeddings, while the decoder guides object queries to focus on class-relevant image features. This design further improves detection accuracy without incurring significant computational overhead. Extensive experiments demonstrate that RT-OVAD consistently outperforms existing state-of-the-art methods across open-vocabulary, zero-shot, and traditional closed-set detection tasks. For instance, on the open-vocabulary aerial detection benchmarks DIOR, DOTA-v2.0, and LAE-80C, RT-OVAD achieves 87.7 AP_50, 53.8 mAP, and 23.7 mAP, respectively, surpassing the previous state-of-the-art (LAE-DINO) by 2.2, 7.0, and 3.5 points. In addition, RT-OVAD achieves an inference speed of 34 FPS on an RTX 4090 GPU, approximately three times faster than LAE-DINO (10 FPS), meeting the real-time detection requirements of diverse applications. The code will be released at https://github.com/GT-Wei/RT-OVAD.

In recent years, the rise of low-cost mini rotary-wing drone technology across diverse sectors has emphasized the crucial role of object detection within drone aerial imagery. Low-cost mini rotary-wing drones come with intrinsic limitations, especially in computational power. Drones come with intrinsic limitations, especially in resource availability. This context underscores an urgent need for solutions that synergize low latency, high precision, and computational efficiency. Previous methodologies have primarily depended on high-resolution images, leading to considerable computational burdens. To enhance the efficiency and accuracy of object detection in drone aerial images, and building on the YOLOv7, we propose the Efficient YOLOv7-Drone. Recognizing the common presence of small objects in aerial imagery, we eliminated the less efficient P5 detection head and incorporated the P2 detection head for increased precision in small object detection. To ensure efficient feature relay from the Backbone to the Neck, channels within the CBS module were optimized. To focus the model more on the foreground and reduce redundant computations, the TGM-CESC module was introduced, achieving the generation of pixel-level constrained sparse convolution masks. Furthermore, to mitigate potential data losses from sparse convolution, we embedded the head context-enhanced method (HCEM). Comprehensive evaluation using the VisDrone and UAVDT datasets demonstrated our model’s efficacy and practical applicability. The Efficient Yolov7-Drone achieved state-of-the-art scores while ensuring real-time detection performance.

End-to-end singing voice synthesis (SVS) model VISinger can achieve better performance than the typical two-stage model with fewer parameters. However, VISinger has several problems: text-to-phase problem, the end-to-end model learns the meaningless mapping of text-to-phase; glitches problem, the harmonic components corresponding to the periodic signal of the voiced segment occurs a sudden change with audible artefacts; low sampling rate, the sampling rate of 24KHz does not meet the application needs of high-fidelity generation with the full-band rate (44.1KHz or higher). In this paper, we propose VISinger 2 to address these issues by integrating the digital signal processing (DSP) methods with VISinger. Specifically, inspired by recent advances in differentiable digital signal processing (DDSP), we incorporate a DSP synthesizer into the decoder to solve the above issues. The DSP synthesizer consists of a harmonic synthesizer and a noise synthesizer to generate periodic and aperiodic signals, respectively, from the latent representation z in VISinger. It supervises the posterior encoder to extract the latent representation without phase information and avoid the prior encoder modelling text-to-phase mapping. To avoid glitch artefacts, the HiFi-GAN is modified to accept the waveforms generated by the DSP synthesizer as a condition to produce the singing voice. Moreover, with the improved waveform decoder, VISinger 2 manages to generate 44.1kHz singing audio with richer expression and better quality. Experiments on OpenCpop corpus show that VISinger 2 outperforms VISinger, CpopSing and RefineSinger in both subjective and objective metrics.

In this paper, we describe our speech generation system for the first Audio Deep Synthesis Detection Challenge (ADD 2022). Firstly, we build an any-to-many voice conversion (VC) system to convert source speech with arbitrary language content into target speaker's fake speech. Then the converted speech generated from VC is post-processed in time-domain to improve the deception ability. The experimental results show that our system has adversarial ability against anti-spoofing detectors with a little compromise in audio quality and speaker similarity. This system ranks top in Track 3.1 in the ADD 2022, showing that our method could also gain good generalization ability against different detectors.

This paper describes our best system and methodology for ADD 2022: The First Audio Deep Synthesis Detection Challenge[1]. The very same system was used for both two rounds of evaluation in Track 3.2 with similar training methodology. The first round of Track 3.2 data is generated from Text-to-Speech(TTS) or voice conversion (VC) algorithms, while the second round of data consists of generated fake audio from other participants in Track 3.1, aming to spoof our systems. Our systems uses a standard 34-layer ResNet [2], with multi-head attention pooling [3] to learn the discriminative embedding for fake audio and spoof detection. We further utilize neural stitching to boost the model's generalization capability in order to perform equally well in different tasks, and more details will be explained in the following sessions. The experiments show that our proposed method outperforms all other systems with 10.1% equal error rate(EER) in Track 3.2.

Following tremendous success in the Generative Adversarial Network(GAN), the GAN-based vocoders have recently shown much faster speed in waveform generation. However, the quality of generated speech is slightly inferior, and the real-time factor (RTF) still can't be satisfied in many devices with limited resources. To address the issues, we propose a new GAN-based vocoder model. Firstly, we introduce the Shuffle-Residual Block into the generator to get a lower RTF. Secondly, we propose a Frequency Transformation Block in the discriminator to capture the correlation between different frequency bins in every frame. To the best of our knowledge, our model achieves the lowest RTF of the GAN-based vocoders under the premise of ensuring the speech quality. In our experiments, our model shows a lower RTF with more than 40% improvement and higher speech quality than MB-MelGAN and HiFi-GAN V2.

This paper presents the details of our system designed for the Task 1 of Multimodal Information Based Speech Processing (MISP) Challenge 2021. The purpose of Task 1 is to leverage both audio and video information to improve the environmental robustness of far-field wake word spotting. In the proposed system, firstly, we take advantage of speech enhancement algorithms such as beamforming and weighted prediction error (WPE) to address the multi-microphone conversational audio. Secondly, several data augmentation techniques are applied to simulate a more realistic far-field scenario. For the video information, the provided region of interest (ROI) is used to obtain visual representation. Then the multi-layer CNN is proposed to learn audio and visual representations, and these representations are fed into our two-branch attention-based network which can be employed for fusion, such as transformer and conformer. The focal loss is used to fine-tune the model and improve the performance significantly. Finally, multiple trained models are integrated by casting vote to achieve our final 0.091 score.

This paper introduces a new open-sourced Mandarin speech corpus, called DiDiSpeech. It consists of about 800 hours of speech data at 48kHz sampling rate from 6000 speakers and the corresponding texts. All speech data in the corpus is recorded in quiet environment and is suitable for various speech processing tasks, such as voice conversion, multi-speaker text-to-speech and automatic speech recognition. We conduct experiments with multiple speech tasks and evaluate the performance, showing that it is promising to use the corpus for both academic research and practical application. The corpus is available at https://outreach.didichuxing.com/research/opendata/.

In this paper we present DELTA, a deep learning based language technology platform. DELTA is an end-to-end platform designed to solve industry level natural language and speech processing problems. It integrates most popular neural network models for training as well as comprehensive deployment tools for production. DELTA aims to provide easy and fast experiences for using, deploying, and developing natural language processing and speech models for both academia and industry use cases. We demonstrate the reliable performance with DELTA on several natural language processing and speech tasks, including text classification, named entity recognition, natural language inference, speech recognition, speaker verification, etc. DELTA has been used for developing several state-of-the-art algorithms for publications and delivering real production to serve millions of users.