Building large-scale generalist pre-trained models for many tasks is becoming an emerging and potential direction in reinforcement learning (RL).Research such as Gato and Multi-Game Decision Transformer have displayed outstanding performance and generalization capabilities on many games and domains.However, there exists a research blank about developing highly capable and generalist models in multi-agent RL (MARL), which can substantially accelerate progress toward general AI.To fill this gap, we propose Multi-Agent multi-Game ENtity TrAnsformer (MA-GENTA) from the entity perspective as orthogonal research to previous time-sequential modeling.Specifically, to deal with different state/observation spaces in different games, we analogize games as languages by aligning one single game to one single language, thus training different "tokenizers" and a shared transformer for various games.The feature inputs are split according to different entities and tokenized in the same continuous space.Then, two types of transformer-based models are proposed as permutationinvariant architectures to deal with various numbers of entities and capture the attention of different entities.MAGENTA is trained on

Existing game AI research mainly focuses on enhancing agents' abilities to win games, but this does not inherently make humans have a better experience when collaborating with these agents. For example, agents may dominate the collaboration and exhibit unintended or detrimental behaviors, leading to poor experiences for their human partners. In other words, most game AI agents are modeled in a "self-centered" manner. In this paper, we propose a "human-centered" modeling scheme for collaborative agents that aims to enhance the experience of humans. Specifically, we model the experience of humans as the goals they expect to achieve during the task. We expect that agents should learn to enhance the extent to which humans achieve these goals while maintaining agents' original abilities (e.g., winning games). To achieve this, we propose the Reinforcement Learning from Human Gain (RLHG) approach. The RLHG approach introduces a "baseline", which corresponds to the extent to which humans primitively achieve their goals, and encourages agents to learn behaviors that can effectively enhance humans in achieving their goals better. We evaluate the RLHG agent in the popular Multi-player Online Battle Arena (MOBA) game, Honor of Kings, by conducting real-world human-agent tests. Both objective performance and subjective preference results show that the RLHG agent provides participants better gaming experience.

This paper describes NIO system for audio-visual diarization and recognition in the Multimodal Information Based Speech Processing (MISP) Challenge 2022. In our system, we proposed combining end-to-end audio-visual neural speaker diarization model and Channel-wise Av-fusion encoder with speaker signature for multi-channel audio-visual speech diarization and recognition. Our system reduces the concatenated minimum permutation character error rate(cpCER) by 34.36% absolute compared to the baseline in track 2.

Triple-level cell NAND flash memory is widely used today due to its higher storage density and capacity. However, with the increase in the storage density, lower reliability results in more read times for flash memory and significantly reduces the read performance. In order to avoid unnecessary read operations, this article proposes a hard decision–soft decoding method called location information-assisted decoding (LIAD) method, which determines the additional information required for decoding by mutual information, and then transmits the required information to correct the log-likelihood ratio (LLR). Different from the conventional LLR correction algorithm, this method does not require additional read operations and correct data. Only using sensing results, our method can reduce uncorrectable error bit rate (UBER) by up to 99%, and the system read latency under SSDsim (Hu et al. 2011) simulation can be reduced by up to 53%.

Although NAND flash memory does a lot of work in effectively using error correcting code (ECC) to reduce uncorrectable bit error rate (UBER). However, if the frame error rate (FER) is not reduced, the lower UBER cannot effectively reduce the read latency of the flash memory system. This phenomenon is especially evident at the end of the flash memory lifetime, where conventional methods significantly reduce the UBER but not to zero, and the remaining error bits are still evenly distributed throughout the flash memory page, resulting in a significant increase in read latency. In this article, an interleaved LDPC decoding scheme is proposed. By re-evaluating the flash memory channel during the decoding process, the codewords in the flash memory page are corrected frame by frame, and the problem of high FER is solved at the end of the flash memory lifetime. Compared with the conventional algorithm, the proposed method can reduce the FER by up to 34%, reduce the average decoding iterations by 63.4%, and reduce the read latency by up to 65%.

In this paper we describe a speaker diarization system that enables localization and identification of all speakers present in a conversation or meeting. We propose a novel systematic approach to tackle several long-standing challenges in speaker diarization tasks: (1) to segment and separate overlapping speech from two speakers; (2) to estimate the number of speakers when participants may enter or leave the conversation at any time; (3) to provide accurate speaker identification on short text-independent utterances; (4) to track down speakers movement during the conversation; (5) to detect speaker change incidence real-time. First, a differential directional microphone array-based approach is exploited to capture the target speakers’ voice in far-field adverse environment. Second, an online speaker-location joint clustering approach is proposed to keep track of speaker location. Third, an instant speaker number detector is developed to trigger the mechanism that separates overlapped speech. The results suggest that our system effectively incorporates spatial information and achieves significant gains.

This paper presents our study of using language branch discriminative information effectively for language recognition. Language branch variability (LBV) method based on factor analysis techniques is proposed. In LBV method, language branch variability factor is obtained by concatenating low-dimensional factors in the language branch variability spaces. Language models are trained within language branches and between languages. Experiments on NIST 2011 Language Recognition Evaluation (LRE) 30s, 10s and 03s tasks show the proposed LBV method provides stable improvement compared to the state-of-art total variability (TV) approach. In 30-second task, it gains relative improvement by 14.6% in equal error rate (EER) and 12.9% in minimum decision cost value (minDCF), and in new metrics of NIST 2011 LRE, it leads to relative improvement of 7.2%-17.7%.

In this paper, we present an effective method to writer identification that is carried out using single Chinese character as script, which is very flexible and very easy to be used in practice. The directional element features are first extracted from the handwriting character scripts, then the dimensions of the features is reduced using PCA in order to cope with the small sample size problem. The most discriminative features are extracted from the reduced feature space using Fisher’s Linear Discriminant Analysis. The Euclidian distance is proposed for classification. Experimental result verified the effectiveness of the proposed method.