To prevent the spread of the COVID-19 pandemic, 2019 has seen unprecedented demand for medical equipment and supplies. However, the problem of waste treatment has not yet been given due attention, i.e., the traditional waste treatment process is done independently, and it is not easy to share the necessary information. Especially during the COVID-19 pandemic, the interaction between parties is minimized to limit infections. To evaluate the current system at medical centers, we also refer to the traditional waste treatment processes of four hospitals in Can Tho and Ho Chi Minh cities (Vietnam). Almost all hospitals are handled independently, lacking any interaction between the stakeholders. In this article, we propose a decentralized blockchain-based system for automating waste treatment processes for medical equipment and supplies after usage among the relevant parties, named Medical-Waste Chain. It consists of four components: medical equipment and supplies, waste centers, recycling plants, and sorting factories. Medical-Waste Chain integrates blockchain-based Hyperledger Fabric technology with decentralized storage of medical equipment and supply information, and securely shares related data with stakeholders. We present the system design, along with the interactions among the stakeholders, to ensure the minimization of medical waste generation. We evaluate the performance of the proposed solution using system-wide timing and latency analysis based on the Hyperledger Caliper engine. Our system is developed based on the hybrid-blockchain system, so it is fully scalable for both on-chain and off-chain-based extensions. Moreover, the participants do not need to pay any fees to use and upgrade the system. To encourage future use of Medical-Waste Chain, we also share a proof-of-concept on our Github repository.
Transfer learning is a common method to improve the performance of the model on a target task via pre-training the model on pretext tasks. Different from the methods using monolingual corpora for pre-training, in this study, we propose a Cyclic Transfer Learning method (CTL) that utilizes both code-switching (CS) and monolingual speech resources as the pretext tasks. Moreover, the model in our approach is always alternately learned among these tasks. This helps our model can improve its performance via maintaining CS features during transferring knowledge. The experiment results on the standard SEAME Mandarin-English CS corpus have shown that our proposed CTL approach achieves the best performance with Mixed Error Rate (MER) of 16.3% on test $_{man}$ , 24.1% on test $_{sge}$ . In comparison to the baseline model that was pre-trained with monolingual data, our CTL method achieves 11.4% and 8.7% relative MER reduction on the test $_{man}$ and test $_{sge}$ sets, respectively. Besides, the CTL approach also outperforms compared to other state-of-the-art methods. The source code of the CTL method can be found at https://github.com/caohongnga/CTL-CSSR .