Nos derniers articles

Computer Vision
Réseau transformer appliqué à la vision
L'objectif de ce billet est de découvrir l’architecture transformer dans les problématiques de vision par ordinateur.
Sebastien Carpentier' width='512' height='512' xlink:href='data:image/png%3bbase64%2ciVBORw0KGgoAAAANSUhEUgAAAEAAAABACAIAAAAlC%2baJAAAACXBIWXMAAAsSAAALEgHS3X78AAAU5ElEQVRo3n3ad9gV5XYF8GMw0puCYG/YsD6CiqiIihWxd7GLiiKINWIXNWIXe48dJRps1x6vN3avXWOJvfcae8vvm8X3OpyPm/3HPHPmzJlZa%2b%2b1yztzGn%2bp7Pbbb7/ttttuvfXWW265xY7tzTffnC2bNm3af7TaTTfdlO0NN9xw4403XnHFFeuss06vXr26du0666yzdurUqUOHDu3bt2/Xrl3Hjh0XXHDBfv36zTvvvJtvvvlhhx02efLkyy677NRTT3XBu%2b%2b%2b%2b9577/3Pyu6///4HHnjgb5X914zWdMRHZ/61Mvu2jb%2b0Gg63VoZAcNcJ2AZ3zP6/VzZ16tShQ4eCPttss80yyyx2/rkyH2et7J8q89Ucc8yx2GKL4bP00kvvvPPOyNx111333XcfGnUCDz74YLA%2bVFlbPuGQnT8J3N5qt9bslspAt28b3IVG0Ps4ZMiQRqMRuBwPfYfKwiQHGRpYCY5YLbfccpttttkpp5wCBPTxaGAVrA%2b1WqH0YGV/q5mfNO64444777wTBzu2t7VaCAS6nWmtFg7Ec/nll%2b%2b1117Q9%2bjRI45vX1mnyuine/fuxLPwwgvb9u7d28f2reZbodh99915QRASAVY081Abe7DVfFvObCEQ6MVKPoRDEVIdfdx/0UUXbb311n379g0mLu/Zsyfo3bp1i4P32Weff6ls77333nTTTRdddFGn4Ql9frLhhhtKIe6LoOu%2bf/jhh/8RhyKk7EyPQIFu5x8Jqc6BeK677roddtgh7rQlcQQWWWSRnXba6bTTTkPv3yoTqHPPPfeggw5ad911V1999ZVWWkk0kiRdunQRQ0ksCOFQ3IzAI4888tDMrCgqNBp3tloiUMpR2aYolYSG3vaCCy6Qu70ri%2bMpe/nllz/88MMvvfTSa6%2b9VoiUKST5GJlzzjnnhBNOOPLIIw855JAttthC1cKZ8Oabbz7Xh6kQCEroE4S2oQiBEoqGcgb9XZUVIZWcLskQSwQcOf/88%2bln4MCBffr0wUGCrrrqqldffXXSuhgCgnDxxRefffbZUvb4449HY%2bLEidtuu62Ide7cWXKj9MQTT0TWwVcIFGtKg3oytBAAvQQhKVHSIASKlqIiOxx83nnnrbzyyjKYEhZffPEpU6YQQ9KG5QrhcNVVV4nYWWedddJJJ0F/3HHHTZgwYbXVVpM2yuuSSy4piYMpQKOfcHj00UejpRKTclo4NKC/5557EoFSjuoltXAoKko%2bEDcQ1LzAAgsg41r1/OEXl3U1HJwJ95577rnddtvtuOOOo0ePHjt27C677CIIVCSFhOjxxx93hQKdPVqzekDqcpouoUCvx6HohxUh1TmARTAyUgege/dIPyJlGQl6NO1qF154IcFoXjJk%2bPDhWvKIESOkgdI0YMAABFzhqKOOeuaZZ%2bihECjQH3vssUKghKWkR0sE6u4vNNLdmlRUMqFwQGChhRZCOFfMLe1H0BR1zTXXcLZZY8011%2bR7Qdh33333339/7teMN9lkE%2b6XP1iBUdK3jr5YiUw9FFTUQuDuViu1KMmQwhomaWr1qurkbbbZZtiwYVIw7hEElcf5GoX4SF9Zq9RK9wMOOEAJ2mOPPTQEEtpvv/1GjRqF0uyzzy4CI0eOTGGJ6IOYqOL%2bfCxaauLwJ4FSiEo2tx2TCoFMY9Jxo402QsBxCUriIJ5%2b%2bulEr5LK3bgfAYgJSb4OHjxY/3KaDrDbbrvNOeecUtm3YCEQ9wd93QqHnFBotBBItpVtXUVxfxOHIiRHuHm99dZz8OSTT4Zv1113PeKII8444wytwNUkg2pDPBtssMG4ceMw0ea22morTVoCEJJWrZGbNUTARQCFqfieX2ZKw049sxv3VuZ%2b2dbzoV5V27YFKvIVr4vDiSeeeOyxxxqV1XtqUXOoSA6IhhPkLsFwM/HArSsbrW2d2b9/f6VMQFTbp556qrj/iZrlYxOTEofGfZUVDnU5hUZyoEyshQMCzuR7qSkUKv0ll1wCro4Ga6Y0BzlbvhIM0Kj%2ba2Xa2dFHH62FCaAICJG4Pfnkk2DV0f%2b9sidmtEIjHKYTCPq2QShVtdSldIYEwRFC510VybCgwgBk5hEQX7msmHAzaaEBvYAAKkpnnnkmGtJaITL5aYjHHHNMARfo%2bPy91dqSKWndyCxVgpBMuKtm9YpU2kKpS%2bZqqpC1mEAGsdFNLXJZrcAJcPO0CCj2AnJxZRrfpEmTENATSEiLMLRCXNxvn6LCoWxZnUk4NO6vLARKEJq6W1NRKp0BmSuvvFIE9DI1pOQ96D5CgCToyo6t0xBA1RwlAlR08MEHy2wDldqK3tNPPx1wT7Za4VC3upBaJBQCiYNt2ziUTKiv3RIBxOje/EzfmShzNb6RxEQFrhq69tprSwNwSUiB4n41V73Cav311zdj05iwaMbxfUA/VdmTM7MZCPy1ZlliJyVCoISiPqWWquo4NFyrXBoccsUUOwRwIzAn6Lt0kjqLRtyvRXC89BUfcx4xxLuFQCIgLPWd8m0pR42yHk0c6gSK1Ue9%2bpThCKdypIoO8bPPPhvf6C8Ykj6V68fSAEpxoHXrMierucYHKwpJL0QABE3Rz1Ot9nTNSkDq2dx4oGZ1DiUZ6kGoZ3MIEIMyOmbMGP5%2b7rnnkogIKK%2bHHnooudO9oiTRrYC33357uBEQB9XWKKUK2VIaTCEQ8TzdxgqlkhVxViNL4zxsqedDPRnSoZsmpRAw7ZAQfKeeemrap0tbu0CvLlG8LA8ZXlc6fWXrfAFZZZVVRGaxxRZzBMQQKI5/pmb5WM%2bKkgaN8nwiw3BTUWoiUG8OiYZc1Ee1VWoB13VdRMpqxuJmMakPmCOcJhkExHJZAmjJZgpDtTJqOaEEiV4hEJdD/GxldQ4lE0KgJQeyMCtxKEFoIlCfL0oy0BIxiICtzqoBE4NvJaWLSGiI7ecxjG5NZrqEUJDTJpUpoyIgrYOprfvrHAqBejFtlOV9eWpXV1Gsnsf1YZuKOBgaESB0NHAQB9dMq3dvl5o8eTIaVjYSxjkiRj%2bWOEsssYQ2PGjQIINQWlhBb1ugl522EmqJQFkd/z%2bZUMpR3F9oGIeIASAFBziuVRwFIcXUbZ5//vkrrrjCksDob3hWNMGVtWTTq1evueeeW1HSRlKCi34C%2brnKnm21sCoc/syBsjpOEAI9aVBP5TJspz2HAG1wPwIGARGQrGYeO3TiIlOnTs0TIZlq9bjMMsvMM888c1bWo0ePbt269e7d2xQkATKBBlxxOfIIPF9ZYZI4lCC0RKA8Q22rolhTbw6BjNlGTv6TA5k0DWpKJHmYnyUux6%2bwwgqGTZ7u2bNn9%2b7du3Tp0rlz5%2b6V4cD91gaSp0k/cf/zM1oJSGkI03MgTynqQWjikDg0dTTpa2toMz9nFoL%2b3MocIe4FKoMbYutGvp933nn79OljDS1rRUMNVawQVr6it6L7OvoXKpspgfTjRtMTr6a2UM/m%2boDE/bbW7EohCeWBnMKv1CiXUkLDgpKP55tvvn79%2bunBhGQNrZfJBzuyXw%2bhPRpD4MUXX6z7/oUZrU6AlQi0dOL6Y5b6S4RCo8zb9RkpQpKgNGDCSQ5oUiYLld5Q7cwDDzxQsUfPjsGO3hD2rZ4wfvx4EVOdkvd0yFkvvfRSpP9CG6tnQtKgNONG/flj00uEEod6QoeACPiozFvyJge4EwEVHQfbPF8w8GWHw1whD%2bWJntdVVXqTOZq0tDZQvfXWW8CB%2b2Kr/XdldgqN1KKoaLqEmh4%2b1vta5rw6gcJBBBxRdhQTmiZoY5lomMy23HJLjkfS/VxHSQXCZa%2b//nplVxAMDvq0KGkOdtQxWa4tOP%2bVV16xDfSXKst%2b%2bOBQ0qAUokbW9k2pXNDPlABwTuBFCdChQ4d27dopL%2br6IpWlSvIrNHzvau7nJwhIfT1LfKzd5IxY6Wt6WR7NC2Pg4gD6yy%2b/bJsgZKdwqLfkRtb2M330XmjcV7OUVOeY74cNG6bIpKL37ds3KdtoNFZccUUxoW9Q8sQzhcvPTd3EI9dpyRjL8WqUCOC/%2bOKLC4ufvPbaay%2b12suVZR%2bHEAiHqKhRHrDMtBY1zdghkCS2TJEA3A9x%2b/btBYH755prLsVHURo8eHD//v3lq6ApMqVOTJs2DQdxUIjgxtxPVNiuXbuSX6ZXtwD0fypDxvbVV1%2bNnNoKaeYEmoRUr6e2CFimmMN4GkozmaJJ9/QgK1yENtaoDCXtbMqUKcUvsp9%2b8lx%2b/vnn1yugFwShUNM0FtUJbStMZ2amInT%2bFgeaDIFUpAShUX9MV6%2bndQJBj4bT7KgtCCj2gPKlmx1bGe/SAFfJ8nXXXdeKEUM%2btnaBCQ0tAh%2b9jOKJTZPWIiSAxueH/KJMyXL7aho%2blt3C5aDf2rq7aOBQCAhCo%2bmZdduuDD0NcECmv7w%2bUkA5Uuc65phjlHaasaBxb6Xd1RVEpcaCfeDAgbwrDrDKChMEqdCMxoxGRjr0XMeNjCfGJ5WKX6zjIlR3d5xrLC3Mi3z3SmUJxZ8EmoJQXhi6Lpdz6mefffbmm28WAnnhDr3SIe4amVnfPUDxq2iXFxGQ4oAiACtnB/3clTmo/JNK3iTkSZkRHfos96AHACUZf1plYpiXdHRBWmj8SaD%2bVrAehND18b333st4DL2ayMdURPSkBSsh%2bQpbfEgzmTdp0iQSh5VgZLkIdOrUSZ1daqmlTNfyNT6SWqDnxX0Z2u3wl21KljveWBlRoSpWGiL%2bjfKUtB6EoOcAYaKHTz/91A5Yb7zxBqAICIKAhoDb5CWkQIupO4GCgKpn66ApY9999zW6qZtGIA07T04T3lQ2W6Dt5L4i4EjeOeDGR4lSfsX9iAmFQbBRfw9V3G/rJIny9ddff/HFF/Tw/fffk5Ad6PmDFgWUfmwR8BEZKkKAe3BAWATkGRp%2baKvw5amJXMqUnoURQO7upi7ipq7GfXnaTiRQ5e0lRzgfjbx6SzlmjaRmUxD8xkV//PHH33//3RaCd999948//pAJl1XmKiaZ5K4zUc1DdpdyBXkS/uSX8SuLD9dPRY4L4YA%2bazEHeTSgE5PUGWfGs2UASyfNepCnGvU3H3nDxRAF%2brfffkPAFvSPPvqIfmhJRdNu4OBvETAXuAdnSy98XJqPpUFenMn%2bDMkhVsYWuDOWOyJoQNM0PpkO6IcHkRc3Jztim0GmdCTXFyW/ajS9gcqrPxd95513/CxFVxp8/PHHb7/9NqB%2bqUpgiL3qjoz7AYrPhAkTXNT9ygsRCFyWnEB0NTeSSNwMSl7F%2bogeB/OLg/bd0S0gSZm34xy3c3ImAzvlwfj0CDRJyBZ65EyXMs9YRuhOdd2MZeqPfUc0LwTc2M20NlXVRe3zE3e6jX3eQiDq10y4Iw8eo3hfOUfc3A5Kpzkh9SfaS0ySnA66WhaD4uC4nzfKG5sSBN/pr5ZOG2%2b8cR7eG7PMNpYdVO5n1KJnSWX6sZ%2b3wo5IAzrmMPdwEA378iw9HyCBEgqEneacwHX3lH/nQ0n6mECSdluWjo4kc5zjo/r2%2beefE0Wj/sopwbWAGj169MiRI82bhpy11lrL6KJxakNKuHKZFzOgG%2bjPqwwgiesrO3Agwwu5sn0XRyBzsiP8h5XTMleCpaQQWByfem0HbednrewKedhD3lJRbfzmm28k5Icfftho%2bybZjJCnsCNGjDCocbNGa1YbM2aMbiombk8w2pYg2Krr0OsP%2bJBy/qES9wPnY9yZtawIcDbEbkTu8HEHSombg6IU6ZdhgSOSP7Yy%2b4cffvj111%2b/%2b%2b47EVAbG/V3ZjHatUoyuqy%2b%2buokLljumj/RjBs3ztAvBwBVRvOPoDTgLBfRME4CnSoEmX07mX6JgQt1MbfwMQ1eNbR1C36xTcdAQHDypKi88PMT7Qj6r776Sl1Br2VJWd4U5LE4Tgq/huVjHvdyko90CRy4eqog8DoCRroISWbTDA0ghoYyCo00hdUggEx5buU0VJ2ZmFAsx7sFkn7iBKwCPT0kj91ff/11oHmd%2b0HPiMELbtdM4JNPPlH1bf0ybbw8SMzbVZo2xlgBkxbo/I2MGpLiIHfLC3ecqQUH9SqzF3OaI2gAB3QmOa51L6GwIwKRTXJA/GXqL7/8AtVPP/0Ep/sCoMBYkEjOGSTk6/%2btDEsXwkfTzRORlF5OIl%2b%2bHz58eF4WQW8o4nXncK1vw4FT47/0YGiI211cx2lAO5JROb3flbFNBPJsQoJCIlltoacLErDMsIa2irLu619Zo/7uW51G11Wy7HJpN8vTHkK3n/9hSQYtwrg/atSoPfbYw0RFJ2QzpbK8wQcuI62LkwQNyLn333%2bfU91FX/eRMu07yM0AKC9EQuUQk8rvlfk2BKjXqg16iySOX74y%2b9MJlCfDmaXSKYU%2bSanA57GP9TgasMItCGjY6g8YKqyO%2b0n%2bEJtGm/bp%2bmntIAosTFB%2b%2b%2b23ujtlwyfxxOfLL7/8rTJwZSrZcD%2bSPnIl0AMGDIA4WyuNNKjpBNLh8ocxPs6gIkf12pMqO6qyiRMnUgs3aw5WtH5vR7swl1MUnmY7yxqVJ/8GYWk9/CIpUxZNWdwPvQWGZPvggw/IBgEH6eTnn3/GITuIISAO%2bhLEQ4cOzZs1ywniQcndG34sjukmoHNhBmaq4FcE4n47CNhKAGHp06dP/uxqbZ7lvInDrK9GjR071lTnt06zZBGxNDXiSU81UAFNq2KCA/1gBYPGxOUiw/cIiJIscoKv1AxYhwwZorjDPX78ePOYOMPcSH%2bh2ogYAWKwtM1zQi434Vj7CII5wkcONoR27NjRCiv/0BWKNdZYQxz22WcfHDAx1TlH%2b7NIQFuxcvH8kyvNSBDgpigOzgNDrMzqVGT5Ab2Co91SBGkA4/oIEI/KA6T6LragKoktS0pRdoO8Q4cPOUtBc0HeWQBt//jKKASaZZddFvTOnTvb9uzZU3u2VrR4t3qUZwK96667oiG5EzTklTL5kywqfxfJi/S8tLVTxhl5Iu8dcb4yxR1kQzyu7Fe8kGWxn2PS8q9Fn4Pe1CllQfRLuP3S7YN%2bYmW%2bMtVRDrgdOnTAwYDUt2/fPpWZlBzPEbOTiqGXm0eEhZxwwMeV6TP/wmTu6Iitb0k3D7ZCTAa6I20T5KBBgyjHEfSIRWmWHnSouP0fNRizjC59IFIAAAAASUVORK5CYII=' /%3e%3c/svg%3e)

Sebastien Carpentier
3 min de lecture

Natural language processing
Révision de contrat grâce au NLP
Dans ce billet de blog, je montre qu’avec les récentes avancées en NLP, les modèles de machine learning peuvent extraire les informations importantes d’un contrat permettant alors un gain de temps considérable ainsi qu’une meilleure gestion des ressources.
Roumaissa Omari' width='200' height='200' xlink:href='data:image/jpeg%3bbase64%2c/9j/2wBDAAYEBQYFBAYGBQYHBwYIChAKCgkJChQODwwQFxQYGBcUFhYaHSUfGhsjHBYWICwgIyYnKSopGR8tMC0oMCUoKSj/2wBDAQcHBwoIChMKChMoGhYaKCgoKCgoKCgoKCgoKCgoKCgoKCgoKCgoKCgoKCgoKCgoKCgoKCgoKCgoKCgoKCgoKCj/wAARCABAAEADASIAAhEBAxEB/8QAGwAAAgIDAQAAAAAAAAAAAAAABgcEBQEDCAL/xAA0EAABAwIEAwYFAgcAAAAAAAABAgMEABEFBhIhBzFBEyJRcYGRCBQyYcEVsSNCUqHR4fH/xAAYAQADAQEAAAAAAAAAAAAAAAADBAUCAf/EACYRAAICAgEDAgcAAAAAAAAAAAECACEDERIEMUEFURMUImGRscH/2gAMAwEAAhEDEQA/AJknL9gTp8hSyzVjceG%2buPBCXHEbKWdwD1t405%2bMcpeB5XKIvdmTXBGaUOaQQStXokH3rmubFeQUtMpKpDoB2F9IOwHmah%2blBsyfFft4lfrc3H6Mcr5uKypD/wDEfX72t6CvSZy%2bzBSVBSNwoHnWEYM8cSRCRZx8q0kjcX67/ajfEckqhYUh1H1pHeuPq8artkVSAZNXE7bIg/huIiXdCwA6Bz8amFtQ3tzoTXrgTErbJGk3H%2bKN2FtvstPI3CgCKw7FTUJjXmLkZsEkCrSG1f0rQEAm9t71cQGhYXAN65z3Msuoy/iKmts5hhQBZao8RT5HQFZI/ZJpHYVPD8hx1JHagatVuXQewB96ZXxQuLZz86tncvwWkBV9gAVAmlfgODYjJR8vBbUXHTcgfyJta6vC9ztWsaY8SBU7ToLMbhnw8gwJOJOPJdbW%2bgX0X71Gua5kCNC7GS8ltaxYJsST6Cq7hhkbGIGLR3JHeZBJdK07kdAPCoec8MmScyzW21uttaVIQto2UhXQ/wCqC4DNZqNISEoXFFmlltLitKVpIJPeQU3HrR7wVybJzvhE9MOQ2mRBWAW1dUqFwf3oNzhg%2bIwkuqlKcdbH0qWdSrfc0wPhQxRUDMGLpCyntGE8kk6rE7f3pzpwjU2iPvEM5dbXvCGfwozLBufkFPoHVog3rVh2TsxLZeWjBZZbY3XcC6fQ7n0pzyeJ2HwZSosyRHZfSQCl1WnmL8%2bRq8w3POFSki7qdR%2booOpI8yKZbpFZdqv4P8i/zDb036nLPGzNmG50zJFkYe28lluP2YWsaSbKve3Tmae/CJOHSOH%2bESo8WMha2AHShABUsd1RP3uK5FkEoairvzFr%2bJv/ANpw/D5m75F6fleeohsrL8Zd%2bV/qR%2bR61P4BFAHiPo2zcdX6iuNjISGXDENwS2m5J6AUssemJXnF5EcLDZuXCpNgT0tR5iUzEY2HPLjusJSCdIWzcpHnelBOnYpLnoLrjIbAspTbR2HmaDksR9FrlK7izIQ3l%2bTpA1KTp9zakjBmyIJC4rzjK/6m1lJ9xRtxSxtMqSnDIy9QQQt03vy5D8%2b1AOkhG/Q0bp10u/eT%2bpfk9eJYYhjE3EdBnSXZCkDSFOq1EDwua9x8axGI%2b0/EmyGX2bdmtDhBTbwqoBrYncmmBUWMJ5DvawlNi10EqQR1HMfmr7KKe1zI%2btBKSWwQRsQb0JBwtx0g82%2b5bzNXOWZhj48re2tJ9udCftDobE6RwjNMxWEfLSWg%2b8kWDlwNXn96BM8YvM/T3VMsBnmAom59Kl5YmB0W1XvvzqFxDNoWhpN1lJt4DalN7Oo6aFRBaVrkPOOEqUVEqUeZ6mstt6oyyepqynRhDbcSbntEpNyPHc1HQ3aMD4m1OgyeR7ypCSVGstqsuxrfHbK9duY3ry61tqHOtiDn/9k=' /%3e%3c/svg%3e)

Roumaissa Omari
4 min de lecture