In this paper, we are interested in the few-shot learning problem. In particular, we focus on a challenging scenario where the number of categories is large and the number of examples per novel category is very limited, e.g. 1, 2, or 3. Motivated by the close relationship between the parameters and the activations in a neural network associated with the same category, we propose a novel method that can adapt a pre-trained neural network to novel categories by directly predicting the parameters from the activations. Zero training is required in adaptation to novel categories, and fast inference is realized by a single forward pass. We evaluate our method by doing few-shot image recognition on the Im-ageNet dataset, which achieves the state-of-the-art classification accuracy on novel categories by a significant margin while keeping comparable performance on the large-scale categories. We also test our method on the MiniImageNet dataset and it strongly outperforms the previous state-ofthe-art methods.

Objective The coronavirus disease 2019 (COVID‐19) has rapidly developed into a pandemic. Increased levels of ferritin due to cytokine storm and secondary hemophagocytic lymphohistiocytosis were found in severe COVID‐19 patients. Therefore, the aim of this study was to determine the role of ferritin in COVID‐19. Methods Studies investigating ferritin in COVID‐19 were collected from PubMed, EMBASE, CNKI, SinoMed, and WANFANG. A meta‐analysis was performed to compare the ferritin level between different patient groups: non‐survivors versus survivors; more severe versus less severe; with comorbidity versus without comorbidity; ICU versus non‐ICU; with mechanical ventilation versus without mechanical ventilation. Results A total of 52 records involving 10 614 COVID‐19‐confirmed patients between December 25, 2019, and June 1, 2020, were included in this meta‐analysis, and 18 studies were included in the qualitative synthesis. The ferritin level was significantly increased in severe patients compared with the level in non‐severe patients [WMD 397.77 (95% CI 306.51‐489.02), P < .001]. Non‐survivors had a significantly higher ferritin level compared with the one in survivors [WMD 677.17 (95% CI 391.01‐963.33), P < .001]. Patients with one or more comorbidities including diabetes, thrombotic complication, and cancer had significantly higher levels of ferritin than those without (P < .01). Severe acute liver injury was significantly associated with high levels of ferritin, and its level was associated with intensive supportive care, including ICU transfer and mechanical ventilation. Conclusions Ferritin was associated with poor prognosis and could predict the worsening of COVID‐19 patients.

Schistosomiasis is a neglected tropical disease that infects 240 million people. With no vaccines and only one drug available, new therapeutic targets are needed. The causative agents, schistosomes, are intravascular flatworm parasites that feed on blood and lay eggs, resulting in pathology. The function of the parasite’s various tissues in successful parasitism are poorly understood, hindering identification of therapeutic targets. Using single-cell RNA sequencing (RNA-seq), we characterize 43,642 cells from the adult schistosome and identify 68 distinct cell populations, including specialized stem cells that maintain the parasite’s blood-digesting gut. These stem cells express the gene hnf4, which is required for gut maintenance, blood feeding, and pathology in vivo. Together, these data provide molecular insights into the organ systems of this important pathogen and identify potential therapeutic targets.