219,244 research outputs found
Burmopsylla Liang, Zhang & Liu 2016
No full text3.188 Genus Burmopsylla Liang, Zhang & Liu, 2016 Burmopsylla Liang, Zhang & Liu, 2016: 484. Type species: Burmopsylla maculata Liang, Zhang & Liu, 2016.Published as part of Guo, Mingxia, Xing, Lida, Wang, Bo, Zhang, Weiwei, Wang, Shuo, Shi, Aimin & Bai, Ming, 2017, A catalogue of Burmite inclusions, pp. 249-379 in Zoological Systematics 42 (3) on page 294, DOI: 10.11865/zs.201715, http://zenodo.org/record/536031
Additional file 1 of “Liu-Liang-Chung” syndrome with multiple congenital anomalies and the distinctive craniofacial features caused by dominant ZEB2 gene gain mutation
No full textAdditional file 1: Supplemental Fig. 1. Imaging findings of the present case. Supplemental Fig. 2. CNVs showing A 22.16 Mb duplication. Supplemental Fig. 3. Fq-PCR showing A 22.16 Mb duplication. Supplemental Table 1. CNVs of cases with Liu-Liang-Chung syndrome
Polyacetylene glucosides from the florets of Carthamus tinctorius and their anti-inflammatory activity
No full textLi, Xin-Rui, Liu, Juan, Peng, Cheng, Zhou, Qin-Mei, Liu, Fei, Guo, Li, Xiong, Liang (2021): Polyacetylene glucosides from the florets of Carthamus tinctorius and their anti-inflammatory activity. Phytochemistry (112770) 187: 1-6, DOI: 10.1016/j.phytochem.2021.112770, URL: http://dx.doi.org/10.1016/j.phytochem.2021.11277
Burmopsylla maculata Liang, Zhang & Liu 2016
No full text249) Burmopsylla maculata Liang, Zhang & Liu, 2016 Burmopsylla maculata Liang, Zhang & Liu, 2016: 485. Type specimen(s). H (♂): EMTG BU-001053 (EMTG). P (♀): EMTG BU-001056, EMTG BU-001087 (EMTG).Published as part of Guo, Mingxia, Xing, Lida, Wang, Bo, Zhang, Weiwei, Wang, Shuo, Shi, Aimin & Bai, Ming, 2017, A catalogue of Burmite inclusions, pp. 249-379 in Zoological Systematics 42 (3) on page 294, DOI: 10.11865/zs.201715, http://zenodo.org/record/536031
Stenopsocus wangi Liang, Li & Liu, 2017, sp. nov.
No full textStenopsocus wangi sp. nov. (Figs. 4–5) Diagnosis. This species is characterized by the dark brown vertex, yellowish pedicel, r-rs without marking and largely sclerotized female subgenital plate. Female. Measurements. Body length 2.78 mm, length from postclypeus to wing tip 4.30 mm. IO: 0.54 mm, d: 0.17 mm, IO/d = 3.18, f1: 0.89 mm, f2: 0.70 mm, FWL: 3.22 mm, FWW: 0.95 mm, HWL: 2.39 mm, HWW: 0.66 mm, t1: 0.34 mm, t2: 0.11 mm. Coloration (in alcohol). Head (Fig. 4 b–d) dark brown except mouthparts. Antennal scape and flagellum dark brown, pedicel yellowish white. Mouthparts yellowish white, with dark brown postclypeus, labral base yellowish, base of maxillary palp dark brown. Prothorax dark purplish brown, meso- and metathorax dark brown. Legs mostly yellowish white, tip of femora and tarsomere 2 brown. Forewing (Fig. 4 e) transparent, wing margin mostly brown, anterior margin of pterostigma yellowish; R dark brown, branch of R and Rs surrounded by a brown marking, a brown stripe present along R 1 in pterostigma. Hind wing (Fig. 4 f) transparent, with a slightly brown marking between costal vein and R. Abdomen dorsally dark purple, genital segments blackish brown. Genitalia (Fig. 5 a) distinctly sclerotized. Epiproct (Fig. 5 b) subtriangular. Paraproct entirely sclerotized, with 22 trichobothrias. Subgenital plate (Fig. 5 c) mostly sclerotized, lateral sclerotized region darker than middle one. Gonapophyses (Fig. 5 d) distinctly sclerotized, external valve subtriangular, dorsal and ventral valve blade-shaped. Male. Unknown. Type material. Holotype female, Laos: Oudomxai, Park Beng, 350 m, 25.III.2016, Xinyue Liu (CAU). Paratype: 1 female, Laos: Oudomxai, near Muang Xai, 880 m, 24.III.2016, Xingyue Liu (CAU). Distribution. Laos (Oudomxai). Etymology. The new species is dedicated to Dr. Guoquan Wang for his kind help during the field trip in Laos. Remarks. This new species appears to be closely related to Stenopsocus eucallus Li & Yang, 1988 (China), Stenopsocus formosanus Banks, 1937 (China) and Stenopsocus signatipennis New, 1978 (Nepal) in having similar forewing marking patterns, but it can be distinguished from the latter three species by the dark brown vertex, yellowish pedicel and largely sclerotized subgenital plate. In S. eucallus, S. formosanus and S. signatipennis, the vertex and pedicel are yellowish, and the middle of female subgenital plate sclerotize indistinctly (Banks 1937; Li 2002; Liang et al., 2015; New 1978). It is necessary to examine more materials to distinguish the latter three species.Published as part of Liang, Feiyang, Li, Fasheng & Liu, Xingyue, 2017, The bark louse family Stenopsocidae (Psocodea: Psocomorpha) newly recorded from Laos, with description of three new species, pp. 589-599 in Zootaxa 4243 (3) on pages 592-595, DOI: 10.11646/zootaxa.4243.3.10, http://zenodo.org/record/40022
Zhengitettix curvispinus Liang, Jiang & Liu 2007
No full text6. Zhengitettix curvispinus Liang, Jiang & Liu, 2007 Specimens examined. 5 ♂ 5 ♀, Fangcheng (Fulong), 21 ° 51 ’N, 107 ° 51 ’E, 500m alt, 21 Jul. 2012, collected by Wei- An DENG. Distribution. China (Guangxi).Published as part of Deng, Wei-An, Zheng, Zhe-Min, Li, Xiao-Dong, Lin, Min-Ping, Wei, Shi-Zhen, Yuan, Bao-Dong & Lin, Li-Liang, 2015, The groundhopper fauna (Orthoptera: Tetrigidae) of Shiwanshan (Guangxi, China) with description of three new species, pp. 151-178 in Zootaxa 3925 (2) on pages 154-155, DOI: 10.11646/zootaxa.3925.2.1, http://zenodo.org/record/24355
Toward scalable and unbiased scene graph generation : active learning and causal inference perspectives
Full text linkAbstract
Scene Graph Generation (SGG) aims to construct structured representations of visual scenes by identifying objects and their pairwise relationships. Despite significant progress, SGG remains challenged by two core issues: the high cost of triplet-level annotations and the persistent biases in relationship prediction. This thesis addresses both challenges through two complementary research threads—label-efficient SGG and bias-mitigated SGG.
In the first thread, we introduce EDAL, a novel active learning framework that integrates evidential uncertainty estimation with diversity-aware sample selection. EDAL enables SGG models to achieve competitive performance using only 10\% of labeled data, dramatically reducing annotation costs while preserving generalization. In the second thread, we investigate the origin of biased predictions in SGG, which are often caused by long-tailed distributions, semantic confusion, and spurious correlations. To this end, we present a series of causal debiasing methods—TsCM, CAModule, and RcSGG—that leverage structural causal modeling, triplet-level logit adjustment, and reverse causal reasoning to systematically mitigate these biases.
Our methods are validated on multiple SGG benchmarks and backbones, achieving state-of-the-art performance on debiasing metrics while preserving overall accuracy. In addition, we highlight the potential of large-scale foundation models in future SGG research, given their success in improving generalization and alleviating annotation burdens in other vision tasks.
By unifying active learning and causal debiasing, this thesis offers a comprehensive framework for building scalable and fair SGG systems, opening new directions for structured visual understanding. Original papers Sun, S., Zhi, S., Heikkilä, J., & Liu, L. (2023). Evidential uncertainty and diversity guided active learning for scene graph generation. In ICLR 2023 : The Eleventh International Conference on Learning Representations. OpenReview.net. https://openreview.net/forum?id=xI1ZTtVOtlz Self-archived version Sun, S., Zhi, S., Liao, Q., Heikkilä, J., & Liu, L. (2023). Unbiased scene graph generation via two-stage causal modeling. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(10), 12562–12580. https://doi.org/10.1109/TPAMI.2023.3285009 https://doi.org/10.1109/TPAMI.2023.3285009 Self-archived version Liu, L., Sun, S., Zhi, S., Shi, F., Liu, Z., Heikkilä, J., & Liu, Y. (2025). A causal adjustment module for debiasing scene graph generation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 47(5), 4024–4043. https://doi.org/10.1109/TPAMI.2025.3537283 https://doi.org/10.1109/TPAMI.2025.3537283 Self-archived version Sun, S., Liu, L., Liu, T., Zhi, S., Cheng, M.-M., Heikkilä, J., & Liu, Y. (2025). A reverse causal framework to mitigate spurious correlations for debiasing scene graph generation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 47(9), 7470–7489. https://doi.org/10.1109/TPAMI.2025.3568644 https://doi.org/10.1109/TPAMI.2025.3568644 Self-archived version Tiivistelmä
Scene Graph Generation (SGG) pyrkii rakentamaan visuaalisista kohtauksista jäsenneltyjä esityksiä tunnistamalla objektit ja niiden väliset parisuhteet. Huolimatta merkittävästä edistyksestä alalla, SGG:tä vaivaavat edelleen kaksi keskeistä haastetta: kolmikkoanotointien korkeat kustannukset sekä pysyvät vinoumat suhdepäätelmissä. Tämä väitöskirja käsittelee molempia ongelmia kahden toisiaan täydentävän tutkimuslinjan kautta: tehokas annotointi (label-efficient SGG) ja vinoumien poistaminen (bias-mitigated SGG).
Ensimmäisessä tutkimuslinjassa esittelemme EDAL:n, uuden aktiiviseen oppimiseen perustuvan kehyksen, joka yhdistää todisteperusteisen epävarmuuden arvioinnin sekä monimuotoisuustietoisen näytevalinnan. EDAL mahdollistaa SGG-mallien kilpailukykyisen suorituskyvyn käyttämällä vain 10\% anotetuista tiedoista, mikä vähentää merkittävästi anotointikustannuksia ilman, että yleistettävyys kärsii. Toisessa tutkimuslinjassa tarkastelemme SGG:n vinoutuneiden ennusteiden alkuperää, jotka johtuvat usein pitkähäntäisistä jakaumista, semanttisesta sekaannuksesta ja näennäiskorrelaatioista. Esittelemme joukon kausaaliperustaisia vinoumanpoistomenetelmiä—TsCM, CAModule ja RcSGG—jotka hyödyntävät rakenteellista kausaalimallinnusta, triplatasoista logit-säätöä ja käänteistä kausaalipäättelyä näiden vinoumien systemaattiseen lievittämiseen.
Menetelmämme on validoitu useilla SGG-vertailuaineistoilla ja selkärankamalleilla, ja ne saavuttavat huipputason tuloksia vinoumametriikoissa samalla säilyttäen kokonaistarkkuuden. Lisäksi tuomme esiin suurten perusmallien (foundation models) potentiaalin tulevassa SGG-tutkimuksessa, erityisesti niiden yleistämiskyvyn ja anotointitarpeen keventämisen ansiosta muissa visuaalisissa tehtävissä.
Yhdistämällä aktiivisen oppimisen ja kausaalisen vinoumanpoiston tämä väitöskirja esittää kokonaisvaltaisen kehyksen skaalautuvien ja reilujen SGG-järjestelmien rakentamiseen, avaten uusia tutkimussuuntia jäsenneltyyn visuaaliseen ymmärrykseen. Osajulkaisut Sun, S., Zhi, S., Heikkilä, J., & Liu, L. (2023). Evidential uncertainty and diversity guided active learning for scene graph generation. In ICLR 2023 : The Eleventh International Conference on Learning Representations. OpenReview.net. https://openreview.net/forum?id=xI1ZTtVOtlz Rinnakkaistallennettu versio Sun, S., Zhi, S., Liao, Q., Heikkilä, J., & Liu, L. (2023). Unbiased scene graph generation via two-stage causal modeling. IEEE Transactions on Pattern Analysis and Machine Intelligence, 45(10), 12562–12580. https://doi.org/10.1109/TPAMI.2023.3285009 https://doi.org/10.1109/TPAMI.2023.3285009 Rinnakkaistallennettu versio Liu, L., Sun, S., Zhi, S., Shi, F., Liu, Z., Heikkilä, J., & Liu, Y. (2025). A causal adjustment module for debiasing scene graph generation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 47(5), 4024–4043. https://doi.org/10.1109/TPAMI.2025.3537283 https://doi.org/10.1109/TPAMI.2025.3537283 Rinnakkaistallennettu versio Sun, S., Liu, L., Liu, T., Zhi, S., Cheng, M.-M., Heikkilä, J., & Liu, Y. (2025). A reverse causal framework to mitigate spurious correlations for debiasing scene graph generation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 47(9), 7470–7489. https://doi.org/10.1109/TPAMI.2025.3568644 https://doi.org/10.1109/TPAMI.2025.3568644 Rinnakkaistallennettu versio Academic dissertation to be presented with the assent of the Doctoral Programme Committee of Information Technology and Electrical Engineering of the University of Oulu for public defence via remote access, on 27 November 2025, at 9 a.m.Abstract
Scene Graph Generation (SGG) aims to construct structured representations of visual scenes by identifying objects and their pairwise relationships. Despite significant progress, SGG remains challenged by two core issues: the high cost of triplet-level annotations and the persistent biases in relationship prediction. This thesis addresses both challenges through two complementary research threads—label-efficient SGG and bias-mitigated SGG.
In the first thread, we introduce EDAL, a novel active learning framework that integrates evidential uncertainty estimation with diversity-aware sample selection. EDAL enables SGG models to achieve competitive performance using only 10\% of labeled data, dramatically reducing annotation costs while preserving generalization. In the second thread, we investigate the origin of biased predictions in SGG, which are often caused by long-tailed distributions, semantic confusion, and spurious correlations. To this end, we present a series of causal debiasing methods—TsCM, CAModule, and RcSGG—that leverage structural causal modeling, triplet-level logit adjustment, and reverse causal reasoning to systematically mitigate these biases.
Our methods are validated on multiple SGG benchmarks and backbones, achieving state-of-the-art performance on debiasing metrics while preserving overall accuracy. In addition, we highlight the potential of large-scale foundation models in future SGG research, given their success in improving generalization and alleviating annotation burdens in other vision tasks.
By unifying active learning and causal debiasing, this thesis offers a comprehensive framework for building scalable and fair SGG systems, opening new directions for structured visual understanding.Tiivistelmä
Scene Graph Generation (SGG) pyrkii rakentamaan visuaalisista kohtauksista jäsenneltyjä esityksiä tunnistamalla objektit ja niiden väliset parisuhteet. Huolimatta merkittävästä edistyksestä alalla, SGG:tä vaivaavat edelleen kaksi keskeistä haastetta: kolmikkoanotointien korkeat kustannukset sekä pysyvät vinoumat suhdepäätelmissä. Tämä väitöskirja käsittelee molempia ongelmia kahden toisiaan täydentävän tutkimuslinjan kautta: tehokas annotointi (label-efficient SGG) ja vinoumien poistaminen (bias-mitigated SGG).
Ensimmäisessä tutkimuslinjassa esittelemme EDAL:n, uuden aktiiviseen oppimiseen perustuvan kehyksen, joka yhdistää todisteperusteisen epävarmuuden arvioinnin sekä monimuotoisuustietoisen näytevalinnan. EDAL mahdollistaa SGG-mallien kilpailukykyisen suorituskyvyn käyttämällä vain 10\% anotetuista tiedoista, mikä vähentää merkittävästi anotointikustannuksia ilman, että yleistettävyys kärsii. Toisessa tutkimuslinjassa tarkastelemme SGG:n vinoutuneiden ennusteiden alkuperää, jotka johtuvat usein pitkähäntäisistä jakaumista, semanttisesta sekaannuksesta ja näennäiskorrelaatioista. Esittelemme joukon kausaaliperustaisia vinoumanpoistomenetelmiä—TsCM, CAModule ja RcSGG—jotka hyödyntävät rakenteellista kausaalimallinnusta, triplatasoista logit-säätöä ja käänteistä kausaalipäättelyä näiden vinoumien systemaattiseen lievittämiseen.
Menetelmämme on validoitu useilla SGG-vertailuaineistoilla ja selkärankamalleilla, ja ne saavuttavat huipputason tuloksia vinoumametriikoissa samalla säilyttäen kokonaistarkkuuden. Lisäksi tuomme esiin suurten perusmallien (foundation models) potentiaalin tulevassa SGG-tutkimuksessa, erityisesti niiden yleistämiskyvyn ja anotointitarpeen keventämisen ansiosta muissa visuaalisissa tehtävissä.
Yhdistämällä aktiivisen oppimisen ja kausaalisen vinoumanpoiston tämä väitöskirja esittää kokonaisvaltaisen kehyksen skaalautuvien ja reilujen SGG-järjestelmien rakentamiseen, avaten uusia tutkimussuuntia jäsenneltyyn visuaaliseen ymmärrykseen
Bisabolane-type sesquiterpenoids: Structural diversity and biological activity
No full textShu, Hong-Zhen, Peng, Cheng, Bu, Lan, Guo, Li, Liu, Fei, Xiong, Liang (2021): Bisabolane-type sesquiterpenoids: Structural diversity and biological activity. Phytochemistry (112927) 192: 1-34, DOI: 10.1016/j.phytochem.2021.112927, URL: http://dx.doi.org/10.1016/j.phytochem.2021.11292
Dipsocus Jie & Liang & Liu 2022, gen. n.
No full textDipsocus gen. n. Type species. Dipsocus fashengi sp. n. Diagnosis. Forewing Sc ending free; Rs and M connected by a crossvein; areola postica nearly trapezoidal; cell m3 normal, not narrowed. Male genitalia: Hypandrium symmetrical, divided into two parts, basal part strongly sclerotized, distal part weakly sclerotized; phallosome ring-like, posteriorly closed. Female genitalia: Epiproct subtriangular; paraproct triangular; subgenital plate with short and distally rounded egg guide, pigmented area slightly sclerotized, roughly T-shaped, stem thread-like and distally bifurcated; gonapophyses: ventral valve slender, dorsal valve broad, pointed at tip, inner margin with strongly sclerotized area at base, external valve with triangular posterior lobe. Distribution. China. Remarks. The new genus can be easily distinguished from the other genera of Thyrsophorini by the male hypandrium divided into two parts, of which the basal part is strongly sclerotized but the distal part is weakly sclerotized; the pigmented area of the female subgenital plate stem is thread-like and distally bifurcated. Etymology. From Greek “ Di- ” (meaning “two”) and “ Psocus ” in reference to the male hypandrium divided into two parts. Gender: Feminine.Published as part of Jie, Lulan, Liang, Feiyang & Liu, Xingyue, 2022, Dipsocus gen. n.: A new bark louse genus of the tribe Thyrsophorini (Psocodea: Psocidae: Psocinae), with description of a new species from China, pp. 94-100 in Zootaxa 5222 (1) on pages 95-96, DOI: 10.11646/zootaxa.5222.1.8, http://zenodo.org/record/745651
Ming Qing Hua yan chuan cheng shi liao liang zhong : "Xian shou zong cheng" yu "Xian shou chuan deng lu" /
No full textZong pai wen ti shi zhong guo fo jiao fa zhan li shi guo cheng dang zhong yi ge nan jie de xian xiang, zong guan zhong guo fo jiao, jing tu yi zong neng fou cheng li zhi jin reng ran ju song fen yun, tian tai yu chan zong you yu you wan zheng de wen xian ji lu, yuan liu chuan cheng xiang dui qing xi. Zhi yu xian shou zong de li shi fa zhan, duo shu fang fu meng long zhi zi, ran er suo xing jin nian xu duo chong yao shi liao xiang ji wen shi, ming qing yi lai xian shou zong fa xi chuan cheng yuan liu ji hu ke wei yun wu kuo qing, zai ci ji chu zhi shang, xue jie dui yu zhong guo fo jiao zong pai fa zhan li shi jin cheng de ren shi yi jing da fu gai xie. "Xian shou zong cheng" yu "xian shou chuan deng lu" shi ming qing yi lai liang bu zhong yao de hua yan chuan cheng shi liao, dui ren shi xian shou zong li shi fa zhan guo cheng ti gong yi ge zhong yao xian suo.宗派問題是中國佛教發展歷史過程當中一個難解的現象,綜觀中國佛教,淨土一宗能否成立至今仍然聚訟紛紜,天台與禪宗由於有完整的文獻記錄,源流傳承相對清晰.至於賢首宗的歷史發展,多屬彷彿朦朧之姿,然而所幸近年許多重要史料相繼問世,明清以來賢首宗法系傳承源流幾乎可謂雲霧廓清,在此基礎之上,學界對於中國佛教宗派發展歷史進程的認識已經大幅改寫."賢首宗乘"與"賢首傳燈錄"是明清以來兩部重要的華嚴傳承史料,對認識賢首宗歷史發展過程提供一個重要線索.Fu lu: fo xin ci ji miao bian da shi bie feng tong gong ta ming deng san zhong.Includes index.附錄: 佛心慈濟妙辨大師別峰同公塔銘等三種Zong pai wen ti shi zhong guo fo jiao fa zhan li shi guo cheng dang zhong yi ge nan jie de xian xiang, zong guan zhong guo fo jiao, jing tu yi zong neng fou cheng li zhi jin reng ran ju song fen yun, tian tai yu chan zong you yu you wan zheng de wen xian ji lu, yuan liu chuan cheng xiang dui qing xi. Zhi yu xian shou zong de li shi fa zhan, duo shu fang fu meng long zhi zi, ran er suo xing jin nian xu duo chong yao shi liao xiang ji wen shi, ming qing yi lai xian shou zong fa xi chuan cheng yuan liu ji hu ke wei yun wu kuo qing, zai ci ji chu zhi shang, xue jie dui yu zhong guo fo jiao zong pai fa zhan li shi jin cheng de ren shi yi jing da fu gai xie. "Xian shou zong cheng" yu "xian shou chuan deng lu" shi ming qing yi lai liang bu zhong yao de hua yan chuan cheng shi liao, dui ren shi xian shou zong li shi fa zhan guo cheng ti gong yi ge zhong yao xian suo.宗派問題是中國佛教發展歷史過程當中一個難解的現象,綜觀中國佛教,淨土一宗能否成立至今仍然聚訟紛紜,天台與禪宗由於有完整的文獻記錄,源流傳承相對清晰.至於賢首宗的歷史發展,多屬彷彿朦朧之姿,然而所幸近年許多重要史料相繼問世,明清以來賢首宗法系傳承源流幾乎可謂雲霧廓清,在此基礎之上,學界對於中國佛教宗派發展歷史進程的認識已經大幅改寫."賢首宗乘"與"賢首傳燈錄"是明清以來兩部重要的華嚴傳承史料,對認識賢首宗歷史發展過程提供一個重要線索.Taiwan resource centre for Chinese studie
- …
